Professional Documents
Culture Documents
Edited by
Peter A Williams
Centre for Water Soluble Polymers, Glyndwr University, Wrexham, UK
Email: williamspa@glyndwr.ac.uk
Glyn O Phillips
Phillips Hydrocolloids Research Ltd, London, UK
Email: phillipsglyn@aol.com
.
View Online
The proceedings of the 17th Gums and Stabilisers for the Food Industry conference held on
2528 June 2013 at the Glyndwr University, Wales, UK.
A catalogue record for this book is available from the British Library
Apart from any fair dealing for the purpose of research or private study for non-commercial
purposes, or criticism or review as permitted under the terms of the UK Copyright, Designs and
.
Patents Act, 1988 and the Copyright and Related Rights Regulations 2003, this publication may
not be reproduced, stored or transmitted, in any form or by any means, without the prior
permission in writing of The Royal Society of Chemistry or the copyright owner, or in the case
of reprographic reproduction only in accordance with the terms of the licences issued by the
Copyright Licensing Agency in the UK, or in accordance with the terms of the licences issued
by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning
reproduction outside the terms stated here should be sent to The Royal Society of Chemistry at
the address printed on this page.
The RSC is not responsible for individual opinions expressed in this work.
Printed in the United Kingdom by CPI Group (UK) Ltd, Croydon, CR0 4YY, UK
Preface
It is a pleasure and a privilege once again to introduce this collection of selected and
reviewed papers from the 17th International Gums and Stabilisers for the Food Industry
Conference held once again at its traditional home at Glyndwr University, Wales,
following its sabbatical to Wageningen, the Netherlands for the 16th Conference. These
volumes have now appeared continuously every two years since their inception in 1982
and have established their position as the most widely referenced publications in the food
hydrocolloids sector and as a partner publication to the journal Food Hydrocolloids.
Significantly proteins have now taken their place tidily alongside polysaccharides within
the food hydrocolloids classification. The papers in Chapters 1 and 4, in particular,
underline this partnership. Peter Wilde outlines clearly the complex field of food proteins.
Subsequent papers show how modification and protein-polysaccharide interactions can
produce new structures and better emulsification properties. Gelatin, of course, has been a
major ingredient in the food industry and the review by Douglas Goff gives us a great deal
of new information about its properties and applications. The papers on emulsions, foams
and films demonstrate major changes in the technologies now used and even in the
nomenclature. The term oxygen cocktails is certainly new to me!
The growth of Chinese activity in this field, particularly with regard to natural bioactive
polysaccharides is reflected not only in the growing input from this country but also in the
increasing amount of collaboration between East and West. The new materials from
biological sources such as Ganoderma atrum, Amorphophallus Muelleri, Pomelo pectin,
.
Okra, Brachystegia Eurycoma etc offer linguistic and structural challenges. However,
they do bridge the traditional food hydrocolloids with these new exotic areas.
Above all it is the health benefits of natural hydrocolloids which are the most appealing
aspect today. The papers on fat replacement, in vitro digestion of dietary fibres, controlling
lipid digestion, controlling human digestion, the synergistic gastric roles of polysaccharide
mixtures, the management of dysphagia etc. show how widespread is the role of food
hydrocolloids in improving human health.
It is a pleasure to note how we all enjoyed the new accommodation and conference
facilities at Glyndwr University. We have become a family after all these years with new
and exciting colleagues joining the old stagers like myself in bringing new subjects and
inspiration into the proceedings. I must thank all concerned, particularly, of course, my
friend Professor Peter Williams, who with our colleague Haydn Hughes undertake most of
the organisation and management of this publication and the conference. I thank also the
View Online
vi Preface
Chairman of the Food Hydrocolloids Trust Dr Graham Sworn, the Trustees and the
Committee Members for their invaluable work in planning and attracting such wonderful
international participation.
Diolch yn fawr i bawb - which is the Welsh for Very many thanks to all
Glyn O. Phillips
Prog. Fd. Nutr. Sci., Gums and Stabilisers for the Food Industry (Eds. G. O. Phillips,
D.J. Wedlock and P. A. Williams). Pergamon Press Ltd, Oxford, Vol 6 (1982).
Gums and Stabilisers for the Food Industry 2 (Eds., G. O. Phillips, D. J. Wedlock and P.
A. Williams), Pergamon Press Ltd., Oxford (1984).
Gums and Stabilisers for the Food Industry 3 (Eds., G. O. Phillips, D. J. Wedlock and
P.A. Williams), Elsevier Applied Science Publishers (1986).
Gums and Stabilisers for the Food Industry 4 (Eds., G. O. Phillips, D. J. Wedlock and P.
A. Williams), IRL Press (1988).
Gums and Stabilisers for the Food Industry 5 (G. O. Phillips, D. J. Wedlock and P. A.
Williams), Oxford University Press Ltd. (1990).
.
Gums and Stabilisers for the Food Industry 6 (eds G.O. Phillips, P.A. Williams and D.J.
Wedlock), Oxford University Press Ltd (1992).
Gums and Stabilisers for the Food Industry 7 (eds G.O. Phillips. P.A. Williams and D.J.
Wedlock), Oxford University Press (1994).
Gums and Stabilisers for the Food Industry 8 (eds G.O. Phillips, P.A. Williams, and D.J.
Wedlock), Oxford University Press (1996).
Gums and Stabilisers for the Food Industry 9 (eds. P.A. Williams and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (1998).
Gums and Stabilisers for the Food Industry 10 (eds P.A. Williams and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (2000).
Gums and Stabilisers for the Food Industry 11 (eds P.A. Williams, P. A. and G.O.
Phillips), Royal Society of Chemistry, Cambridge UK (2002).
Gums and Stabilisers for the Food Industry 12 (eds P.A. Williams and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (2004).
View Online
viii Conferences
Gums and Stabilisers for the Food Industry 13 (eds P.A. Williams, and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (2006).
Gums and Stabilisers for the Food Industry 14 (eds P.A. Williams, and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (2008).
Gums and Stabilisers for the Food Industry 15 (eds P.A. Williams, and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK (2009).
Gums and Stabilisers for the Food Industry 16 (eds P.A. Williams, and G.O. Phillips),
Royal Society of Chemistry, Cambridge UK, (2012)
.
Acknowledgements
The Food Hydrocolloid Trust is indebted to the members of the conference organising
committee for their efforts in arranging this conference.
x Acknowledgements
Sponsors
The Food Hydrocolloid Trust is grateful to the following Companies and Organisations for
sponsoring the Conference.
Biopolymer Solutions
CP Kelco
Doehler Group
Du Pont
Elsevier
FMC Biopolymers
Kerry Ingredients
Mars GmbH
.
Setaram Instrumentation
Stable Microsystems
Wyatt Technologies
Contents
CHAPTER 1 1
Milk proteins 11
H.D. Goff, University of Guelph, Guelph, Canada
CHAPTER 2 71
xii Contents
Germany
CHAPTER 3 165
RHEOLOGICAL PROPERTIES
Thickening properties of corn fiber gum with other carbohydrate polymers 167
M. Yadav, F. Zhang, T. Luan, L. Wu and H. Zhang, Eastern Regional Research
Center, Wyndmoor, USA, Shanghai Jiao Tong University, Shanghai, China
Contents xiii
Phase separation and gel formation in kinetically-trapped guar gum / acid 205
milk gels
A. Rohart and C. Michon, AgroParis Tech, Massy, France, INRA, Massy, France,
CNAM. Paris, France
Compression test of food gels on an artificial tongue and its comparison with 214
sensory tests
S. Ishinara, M. Isono, S. Nakao, M. Nakauma, T. Funami, K. Hori, T. Ono, K.
Kohyama and K. Nishinari, San-Ei Gen F.F.I. Inc. , Osaka, Japan, Niigata
University Graduate School of medical and Dental sciences, Niigata, Japan,
Osaka Univerisity School of Dentistry, Osaka, Japan, National Food Research
Institute, Tsukuba, Japan
CHAPTER 4 221
xiv Contents
CHAPTER 5 287
HEALTH ASPECTS
Rheological behaviour of maize -glucan and its application as a fat replacer 318
in baked goods
G.O. Sampson, A.Y. Tettyeh and J.H. Oldham, Kwame Nkrumah University of
Science and Technology, Kumasi, Ghana
.
Effects of soluble dietary fibres on glucose mobility and starch hydrolysis 328
during in vitro digestion
H. Fabek and H.D. Goff, University of Guelph, Guelph, Canada
Interactions between polymeric surfactants and bile salts: new routes for 334
controlling lipid digestion or oil-in-water emulsions
A. Torcello-Gomez, J. Maldonado-Valderrama, A.B. Jodat-Reyes and T.J. Foster,
University of Nottingham, Nottingham, UK, University of Granada, Granada,
Spain
Interactions between hydrocolloids and bile salts during human digestion of 342
emulsions
C. Fernandez Fraguas, N.C. Woodward, A.P. Gunning and P.J. Wilde, Institute
of Food research, Norwich, UK
Comparison of two tests used for the classification of food thickeners in the 359
management of dysphagia
C. de Saint-Aubert, G. Sworn and J. Kayasita, DuPont Nutrition Biosciences ApS,
Braband, Denmark, DuPont, Danisco France SAS, Paris, France, Prefectural
University of Hiroshima, Hiroshima City, Japan
View Online
Contents xv
Peter J. Wilde
Institute of Food Research, Norwich Research Park, Norwich, NR4 7UA, UK.
1 INTRODUCTION
Protein is one of the three vital macronutrients required for growth and development and
maintenance of a whole range of physiological functions. Proteins come from a variety of
plant and animal sources, including meat, fish, cereals, pulses and dairy. Their amino acid
composition is the primary nutritional value. The amino acids are utilised for the synthesis
of proteins and used as a nitrogen source for the synthesis of other biologically important
molecules. There are 20 amino acids, 9 of which are deemed essential or indispensable for
human health, as they cannot be synthesised in vivo, so require a dietary source.1
.
Protein sources are currently a topical issue as feeding the rapidly growing global
population sufficient protein is fast becoming a global issue of food security.2 Hence
alternative, more effective production of rich sources of dietary protein are being
considered such as mycoprotein, cultured or in-vitro meat and insect protein.
As well as being an essential nutrient, proteins have also been traditionally used to confer
functionality to food products, primarily to manipulate the texture of foods3 to aid their
acceptability or to mimic the texture of other foods (e.g. fat replacement). The unique
structural polymorphisms of proteins and their interactions on a range of length-scales
allow this manipulation of structure and texture to take place in order to improve the
control of food functionality. There is also a growing interest in the bioactivity of certain
proteins and peptides and their role in physiological processes such as appetite, blood
pressure, colitis and allergy.4 However, in this article we will concentrate on the generic
role which proteins have as functional components to control the structural and textural
properties of food.
2. PROTEIN STRUCTURE
and influence the overall protein structure and functionality.3 The sequence of amino acids
in a protein is known as the Primary structure. The various molecular properties of the
different amino acid side chains, and the chemical conditions they encounter, can induce
turns and bends in the polypeptide backbone of the protein. This leads to the formation of
helices, sheets and coils known as Secondary structure. The various secondary structures
my then interact to form the overall structure of the protein molecule, perhaps to form
active sites in enzymes, or hydrophobic pockets to bind certain molecules. These structures
can be covalently stabilised by internal disulphide bridges between cysteine amino acids.
This level of assembly is the Tertiary structure. Finally, Quaternary structure is the
complexation of two or more individual protein molecules or polypetides to form a larger
protein complex with specific functionality such as membrane channels or microtubules.
3. PROTEINS IN FOOD
Proteins play a huge variety of essential physiological roles within plants and animals,
hence dietary protein is an essential nutrient. They also contribute significantly to the
texture of natural, unprocessed raw materials to differing degrees, depending on their
physiological role. In some forms, the intrinsic structure and form of the protein plays the
major structural role, in others, some form of processing is required to utilise protein
functionality.
In most meat and fish products, the primary form of protein is actin and myosin in
myofibrils and muscle fibres. This fibrous tissue gives the food its unique structure, and the
textural properties can vary between source. They respond to cooking in different ways due
to the denaturation of the proteins and water loss, causing a major changes in texture.5
.
Other important meat proteins include those from connective tissues, mainly collagen,
which imparts strengthening and structural stability to tissues. Collagen forms regular
helices, which combine to form structured fibres responsible for their structural integrity
and physical properties.6 Connective tissue tends to be very tough and not consumed
widely without extensive cooking. However, collagen can be processed using acid, alkali
or enzymatic treatments to produce gelatin which have excellent gelling properties due to
its triple-helix structure, and strength of intermolecular interactions to form a strong elastic
network.7
All dairy products are derived from milk, where the main role of the protein is to act as a
bioavailable source of protein, stabilise poorly soluble nutrients such as vitamins, minerals
and fat.8 Simple processing such as removing water by creaming and concentrating the
milk, allows the proteins and fat globules to interact forming more viscous, high fat,
cream-based products. Interestingly, the structure and composition of casein micelles in
milk, promote extensive aggregation when they enter the stomach, where the acid and
proteolytic enzymes allow coagulation to form viscoelastic structures which slow down
their passage through the GI tract, allowing a higher degree of digestion and hence nutrient
availability for the infant. It is no coincidence therefore that acidic and proteolytic
treatments are used to process milk to form more textured products such as yoghurt and
View Online
cheese, where the milk proteins form a major role in the structure and texture of the
product.9
Proteins generally do not play a major role in the texture of the product when consumed in
its natural state (cereals, nuts, legumes, fruit and vegetables), as the structure tends to be
dominated by the cellular structure of the tissues. However, when processed, the protein
may readily interact with other components to influence the texture of the product. One
example is bread; although the protein content of bread flour is only around 12-14%, when
hydrated and mixed, the gluten proteins interact strongly to form a network which adds
elasticity to the starch rich matrix, giving bread its unique properties. Other plant sources
of protein have to be processed in order to make the protein available for use as a food
structural ingredient. The most widely used protein source is soya, which has a protein
content of up to 40 % by weight. The protein can be extracted, after oil extraction if
required, to be used as a protein source in a wide range of foods.10 Other products can be
produced, such as tofu, which is formed by precipitating soy milk, and is typically
composed of 10% protein.11 Another popular source is mycoprotein,12 which is not a
wholly natural form, but which is produced by fermentation growth of Fusarium
microfungi which is separated and processed to form a range of structured protein-based
food products.
We will discuss the main structural features for which proteins are most commonly used to
create and stabilise in foods. These are gels, foams and emulsions and finally, a section
.
covering the more complex structures and textures such as fibrous materials.
4.1 Gels
As proteins are amphipathic macromolecules, they are generally able to form gels.
However, their diversity of structure dictates that the microstructures and mechanisms of
gelation can vary significantly from extended interacting networks, to condensed
particulate dispersions.13, 14 The formation of a gel requires the creation of an interacting
network between individual molecules or stable aggregates15 (primary element). The
controlling factors are the concentration of the primary element, and the strength of the
interaction between them.
The morphology of the primary element is important to control the gelation concentration,
as rods and fibres can interact over longer length scales than individual molecules or small
globular aggregates.13 For example gelatin forms helical, rod shaped molecules, which
unfolds into long strands when heated, allowing interactions with neighbouring molecules
to form across extended length-scales, which, when cooled arrest the structure, forming an
elastic gel at low protein concentrations ca. 1wt%.15 Whereas -lactoglobulin forms small
globular aggregates, which do not unfold significantly upon heating due to their compact
tertiary structure and low molecular weight, and therefore requires a much higher
concentration (ca. 10wt%) to gel properly.16
The interactions between the primary elements are also critical for the formation and
strength of the gel network. Interactions can be non-covalent through electrostatic,
hydrophobic and hydrogen bond interactions, or they can be covalent, through disulphide
View Online
bridges or enzymatic crosslinking reactions. Therefore pH and ionic strength can play a
huge role in determining the strength of interactions and also control the morphology of the
aggregates.14 As proteins are poly-electrolytes, containing amino acid residues with a range
of pKa values, varying pH will change the charge distribution over the molecule,
controlling the electrostatic interactions between the proteins. The interactions tend to be
strongest at the iso-electric point of the protein, where the overall net charge is close to
zero. This is often used as a method to promote aggregation and gelation in milk
processing17, 18 and also to induce phase separation of proteins, for example in cheese
production to separate the casein rich curds from the whey, in order to concentrate the
protein into a gel prior to processing and maturation. Net repulsive electrostatic
interactions tend to favour the formation of fine-stranded structures, whereas particulate
gels are generally formed when there is little or no repulsion.3 This may occur near the
isoelectric point of the protein, or where the high ionic strength screens out the repulsive
forces. pH can also be used to manipulate the primary elements. There has been a lot of
interest recently in protein nano-fibrils, particularly milk proteins, which are made by
hydrolysing the protein at low pH and raising the pH, and the peptides remaining
spontaneously form fibrils and tubules.19-21 These structures can then be further processed
to form gels with dramatically different functional properties compared to the native, intact
protein.
Enhancing the interactions between proteins through processing can enhance the gelling
properties. For example, transglutaminase catalyses the formation of covalent bonds
between lysine and glutamine residues on proteins. This can dramatically strengthen the
interactions between proteins in a gel and is used in the food industry to increase gel
strength in reconstituted meat products where the gelation properties of the intrinsic
proteins are weak.22 Shear can also enhance interactions, the classic example being bread
dough, where the shear forces encountered during mixing and kneading allow the various
gluten proteins to come into contact and interact through covalent, hydrophobic and
.
hydrogen bonding. This allows the gluten to form an elastic network within the starch
matrix to help stabilise the dough structure and gas cells before and during baking.
Therefore the diverse structure and molecular properties of proteins and their ability to
interact, can be exploited to produce gels with varied rheological and textural behaviours to
suit particular foods.
Proteins are widely used to stabilise foams and emulsions in both food and non-food
applications.23 Creating a foam or emulsion involves the formation of many small bubbles
or droplets. This results in a huge increase in the interfacial area of the system. The energy
barrier to this surface expansion is the surface free energy or surface tension. The surface
or interfacial tension is defined as the force per unit length acting on a surface. Therefore a
lower surface tension induced by the adsorption of stabilising molecules will require less
energy to produce a specific increase in surface area. In order to occupy and stabilise this
surface requires an amphiphilic molecule, containing both hydrophobic and hydrophilic
parts. These can be small molecular weight surfactants and emulsifiers, or amphiphilic
polymers such as proteins.3, 23 Proteins are composed of amino acids with varying charge
and hydrophobicity, and often, because of their physiological role, they often have
hydrophobic regions which are utilised in their native role to bind lipophilic ligands or
interact with membranes and surfaces. There are a range of these examples in nature such
as the proteins in the milk fat globule membrane which help to stabilise milk fat globules,
View Online
oleosins in plants which stabilise oil bodies in plant cells and fungal hydrophobins which
adsorb to hydrophobic surfaces to change their wetting properties.
The speed at which proteins can adsorb to a surface and reduce the interfacial tension is
critical for the initial formation of foams and emulsions.24 The rate of change of interfacial
tension is heavily influenced by the molecular weight and hydrophobicity of the protein.
Smaller molecular weight proteins will diffuse more quickly to the interface and often
result in increased functionality.25 Hydrophobic proteins, if soluble, will lower the
interfacial tension more effectively, and improve functionality.26 Increasing hydrophobicity
and hence functionality has been achieved by thermal treatment,27 and chemical
modification such as succinylation.28 There appears to be a fairly straightforward link
between rate of change of interfacial tension and the ability to create protein foams and
emulsions. However, maintaining long term stability requires very different surface
properties.
Once formed, the adsorbed layer surrounding the bubbles or droplets should act to prevent
destabilisation of the foam or emulsion through coalescence, flocculation or creaming.
Proteins can use different mechanisms to maintain stability. Electrostatic or steric
hindrance can prevent close approach of the droplets in order to maintain a stable,
dispersed system. At high concentrations, or under shear, the droplets will come into close
contact, and a different stability mechanism is required. Short range stability mechanisms
involve either the formation of a viscoelastic network on the surface or a highly mobile
Gibbs-Marangoni mechanism normally favoured by surfactants. Proteins tend to have zero
or extremely slow diffusion within an adsorbed layer, therefore proteins which can
stabilise foams and emulsions in the short range tend to be those which can form a
viscoelastic film on the surface of the bubbles or drops. As the droplets come into contact,
their shape changes, leading an expansion of the surface area. A stable visco-elastic layer
will expand with the droplet and as long as the surface layer does not rupture, or is
expanded beyond its limit, then the bubble or droplet should remain stable. The mechanical
.
properties of protein stabilised interfaces are therefore important for stabilising foams and
emulsions,29, 30 and hence interfacial rheology has been a popular method for studying the
surface properties of proteins. In addition to protein structure, the interactions between
proteins are vital for the development of an elastic interface,31 in a similar way that they
are important for their gelling properties.3 The interactions can be in the form of
electrostatic and hydrophobic interactions, hydrogen and covalent bonds. The importance
of electrostatic interactions was demonstrated by the observation that many proteins
display maximal surface elasticity close to their isoelectric point.26 Covalent bonds have
also been observed between adsorbed proteins.32 Enhancing surface rheological properties
(e.g. by altering pH) have been found to coincide with increased foam and emulsion
stability,26, 33 further supporting the importance of these surface properties for functional
performance.
Therefore, to successfully create and stabilise foams and emulsions, a protein has simply to
be soluble, hydrophobic and form a strong elastic interface. This is relatively easy to
achieve in a model system, but not in a real food product, which may have passed through
a range of processes, not optimised for protein functionality. Therefore, the optimal
conditions for foam and emulsion stability in foods is at best, a compromise between
solubility and hydrophobicity and surface aggregation.
A major source of protein in the human diet is meat and fish, which have very specific
textures and structures, largely due to the structure and interactions of the contractile
View Online
proteins (actin and myosin) and the myofibril structures within muscle fibres. For various
reasons, including food security, cost of livestock production, health and dietary choices,
non-meat protein is becoming the healthier, more sustainable choice. However, many
consumers appear to prefer the meat-like structure to their protein, as evidenced by the
existence of the ubiquitous veggie-burger. Hence, methods were developed to process
vegetable proteins, initially using soya, to texture the proteins to resemble that of meat.
The most common methods is extrusion cooking,34 and since the 1970s there have been a
large number of patents describing various methods to induce structure in vegetable
proteins to produce meat substitutes. The specific shear and elevated temperature (<140C)
encountered in the extruder can induce electrostatic and covalent bonds to develop between
proteins.34 The design and temperature profile of the die can induce multilayered and
fibrous textures,35 and the level of expansion allowed to occur after the product emerges
from the dye can control the product density and structure. This has led to the development
of a whole range of meat analogue products based not only on vegetable protein, but also
proteins specifically produced by fungal fermentation (mycoproteins). However, the key
element is the design of the thermal processing used to convert largely globular proteins
into fibrous aggregates.
Phase separation of proteins and polymers can be used to induce specific rheological
properties.36 Specifically, the incompatibility between polymers, whether molecular or
thermodynamic can induce them to phase separate under certain conditions, influenced by
temperature, pH, ionic strength, concentration etc. One of the main applications in food is
to replace fat in emulsified foods. By creating a polymer mixture in which one phase forms
small droplets, similar to oil droplets in an emulsion,37 the rheological behaviour of an
emulsion can be created, but in the absence of fat. The phase behaviour and processing
conditions here are critical, to allow the formation of droplets of the appropriate size and
physical properties, in order to mimic the sensory properties (in terms of texture) of an oil
in water emulsion.38
.
As discussed earlier, with the exception of rod-shaped proteins like gelatin, the gelling
properties of most proteins is based on largely spherical or fractal aggregates, which means
that high concentrations of protein are required to form a gel. There has been increasing
research interest in modifying proteins to form linear aggregates or micro- / nano-fibrils.19
Although much of the research on nano-fibrils is in the amyloid disease area,39 this
research has been adapted to potentially exploit the novel properties of these structures in
food systems. Various approaches have been attempted, but most successful approaches
involve heating proteins, often whey proteins or albumins, which have been hydrolysed at
low pH.19 This then instigates a self assembly of peptides into rod-like fibrils, the structure
and size of which will depend on the pH, temperature, shear and ionic strength.20 However,
these fibrils are capable of forming gels at protein concentrations of less than 0.1% for -
lactoglobulin,21 which is more than an order of magnitude lower than is often observed.40
Electrospinning techniques have been used for producing ultrathin protein and polymer
fibres. There has been interest recently for a range of medical applications such as wound
dressing, tissue scaffolds 41 and drug delivery applications.42 The potential applications in
food systems have not yet been fully explored yet, but there is significant interest in this
method for producing robust, sub-micron fibrils for a variety of food applications including
novel structures, encapsulation and release as well as processing aids such as active
filtration and novel packaging.43
View Online
5 CONCLUSIONS
To summarise, proteins have been utilised for many years to illicit a wide range of
structures into foods, primarily to impart desirable taste and texture for consumers. There
is a wealth of knowledge regarding the various processing methodologies which can utilise
the unique structure, interactions and physicochemical properties of proteins to manipulate
their behaviour. However, as we look to the future, the key global challenges of food
security, and a healthy, ageing population are driving research into new directions. The
issues of providing the worlds population with sufficient, sustainable sources of protein2
will require novel approaches to process alternative protein sources into acceptable food
products. The importance of protein and peptide bioactivity is beginning to gain
importance,4 together with diet related conditions associated with protein malabsorption
such as sarcopaenia,.44 Further interest is also developing in the use of protein structures to
deliver essential micronutrients and drugs.45, 46 Therefore research is needed to ensure that
food proteins are available in appropriate forms to optimise the impact on diet and health
and that they are in a form which is acceptable to the consumer to add value and
functionality to food proteins and address the global food and health issues.
6 REFERENCES
6. L. Pauling and R. B. Corey, Proc. Natl. Acad. Sci. U. S. A., 1951, 37, 272-281.
7. M. C. Gomez-Guillen, B. Gimenez, M. E. Lopez-Caballero and M. P. Montero, Food
Hydrocolloids, 2011, 25, 1813-1827.
8. P. F. Fox, Proc. Nutr. Soc., 1978, 37, 247-257.
9. P. F. Fox, Int. J. Dairy Technol., 2001, 54, 41-55.
10. D. Fukushima, in Handbook of Food Proteins, eds. G. O. Phillips and P. A. Williams,
Woodhead Publ Ltd, Cambridge, Editon edn., 2011, pp. 210-232.
11. H. L. Wang, Journal of the American Oil Chemists Society, 1984, 61, 528-534.
12. T. J. A. Finnigan, in Handbook of Food Proteins, eds. G. O. Phillips and P. A.
Williams, Woodhead Publ Ltd, Cambridge, Editon edn., 2011, pp. 335-352.
13. R. Mezzenga and P. Fischer, Reports on Progress in Physics, 2013, 76.
14. E. van der Linden and E. A. Foegeding, Gelation: Principles, Models and
Applications to Proteins, Elsevier Academic Press Inc, San Diego, 2009.
15. P. A. Williams and G. O. Phillips, The use of hydrocolloids to improve food texture,
Woodhead Publishing Ltd, Cambridge, UK, 2003.
16. J. Gezimati, L. K. Creamer and H. Singh, J. Agric. Food Chem., 1997, 45, 1130-1136.
17. D. S. Horne, Current Opinion in Colloid & Interface Science, 2002, 7, 456-461.
18. T. van Vliet, C. M. M. Lakemond and R. W. Visschers, Current Opinion in Colloid &
Interface Science, 2004, 9, 298-304.
19. E. van der Linden, Food Hydrocolloids, 2012, 26, 421-426.
20. D. Oboroceanu, L. Z. Wang, A. Kroes-Nijboer, A. Brodkorb, P. Venema, E. Magner
and M. A. E. Auty, Int. Dairy J., 2011, 21, 823-830.
View Online
21. C. Veerman, L. M. C. Sagis and E. van der Linden, Macromol. Biosci., 2003, 3, 243-
247.
22. J. Buchert, D. E. Cura, H. Ma, C. Gasparetti, E. Monogioudi, G. Faccio, M. Mattinen,
H. Boer, R. Partanen, E. Selinheimo, R. Lantto and K. Kruus, in Annual Review of
Food Science and Technology, Vol 1, eds. M. P. Doyle and T. R. Klaenhammer,
Annual Reviews, Palo Alto, Editon edn., 2010, vol. 1, pp. 113-138.
23. P. Wilde, A. Mackie, F. Husband, P. Gunning and V. Morris, Adv. Colloid Interface
Sci., 2004, 108, 63-71.
24. S. Nakai and E. Lichan, Crit. Rev. Food Sci. Nutr., 1993, 33, 477-499.
25. L. P. Grunden, D. V. Vadehra and R. C. Baker, J.Food Sci., 1974, 39, 841.
26. J. R. Mitchell, In Developements in Food Proteins - 4 (Ed B.J.F.Hudson) Elsevier
London, 1986.
27. M. Corredig and D. G. Dalgleish, J. Dairy Res., 1996, 63, 441-449.
28. O. Toledano and S. Magdassi, J. Colloid Interface Sci., 1998, 200, 235-240.
29. P. J. Halling, Crc Critical Reviews in Food Science and Nutrition., 1981, 15, 155-203.
30. A. Prins, Colloids and Surfaces A-Physicochemical and Engineering Aspects, 1999,
149, 467-473.
31. S. Damodaran, Adv.Food Nutr.Res., 1990, 34, 1-79.
32. E. Dickinson and Y. Matsumura, Int.J.Biol.Macromol., 1991, 13, 26-30.
33. K. Maeda, S. Yokoi, K. Kamada and M. Kamimura, Am.Soc.Brew.Chem.J., 1991.
34. D. A. Ledward and R. F. Tester, Trends Food Sci. Technol., 1994, 5, 117-120.
35. J. C. Cheftel, M. Kitagawa and C. Queguiner, Food Rev. Int., 1992, 8, 235-275.
36. V. Tolstoguzov, Biotechnol. Adv., 2006, 24, 626-628.
37. M. A. K. Williams, D. Fabri, C. D. Hubbard, L. Lundin, T. J. Foster, A. H. Clark, I. T.
Norton, N. Loren and A. M. Hermansson, Langmuir, 2001, 17, 3412-3418.
38. B. J. D. Le Reverend, I. T. Norton, P. W. Cox and F. Spyropoulos, Current Opinion in
Colloid & Interface Science, 2010, 15, 84-89.
.
H. Douglas Goff
1 INTRODUCTION
Cows milk, the principal source of milk for human food, is the source of two main
categories of proteins - the caseins and the serum or whey proteins. These have very
different compositional, structural and functional properties. This brief chapter will review
these properties for each category of milk proteins. It will also discuss ingredient products
that are derived from milk containing varying mid to high levels of these proteins and their
12/04/2014 12:04:43.
2.1 Caseins
The casein proteins are the ones found within the colloidal casein micelle in milk after
secretion, which precipitates from milk at its isoelectric point of pH 4.6. Caseins show
minimal secondary and tertiary structure and do not denature upon heating. They are
generally hydrophobic and fairly highly negatively charged, which keeps them in colloidal
View Online
suspension. They readily associate with each other and self-associate. There are four
principal components, named as Ds1-, Ds2-, E- and N-casein, which comprise 41%, 11%,
33% and 12% of the weight of caseins, respectively3. The primary sequences of all of the
casein proteins have been determined1. The D- and E-caseins are phosphorylated and vary
from 23-25 kDa, while the N-casein, which is glycosylated, has a molar mass of 19.5 kDa3.
Ds1-casein contains a high net negative charge and a high phosphate content. E-casein is
quite amphiphilic, with a polar head group of 44 residues that contains very little proline
and a high negative charge and an apolar tail of 165 residues that contains a high content of
proline and a nearly-neutral charge3. N-casein has two cysteine residues that can form
intermolecular disulphide bonds. Its peptide bond between residues 105 and 106 is readily
cleaved by proteolytic enzymes, notable chymosin or rennet, and this is the principal milk
clotting reaction in the formation of a milk gel as a precursor to curd formation in
cheesemaking.
The unique properties of the individual casein molecules described above allows
them to easily aggregate into the colloidal casein micelle. Its structure has been the subject
of considerable research4,5. Ds1- and Ds2-casein provide the main structural network, to
which E-casein is attached through hydrophobic interactions. The phosphate groups of
these proteins react extensively with ionic calcium, so that the colloidal calcium phosphate
is instrumental in bonding the proteins together. The milk salt equilibrium between
colloidal and soluble salts is thus critical for micellar protein interactions, and is often
manipulated industrially to either enhance or weaken these interactions. As hydrophobic
bonds get weaker at low temperatures, some E-casein can dissociate from the micelle into
12/04/2014 12:04:43.
the serum phase3. N-casein is mostly found at the surface of the micelle with its residues
106-169 forming what has often been referred to as the hairy layer, providing steric
stabilization to the micelle. When this caseinomacropeptide (or glycomacropeptide) is
cleaved by rennet, the micelles easily aggregate5.
There are two principal serum proteins, E-lactoglobulin, an 18 kDa protein that comprises
about 52% of the total serum protein, and D-lactalbumin, a 14 kDa protein that comprises
about 20% of the total serum protein3. The balance is found in several minor proteins: the
immunoglobulins, bovine serum albumin and peptides from incomplete synthesis of the
caseins. The serum proteins are soluble, globular proteins with a high proportion of D-helix
and E-sheet conformation. Their primary sequences and secondary and tertiary
conformations have been determined1. They will denature with heat, typically >70C
although the precise temperature is dependent on their environment, especially pH. When
denatured in milk, they can react with the surface of the casein micelle and can also form
soluble serum protein aggregates. E-lactoglobulin contains a single sulfhydryl group within
its native globular structure that reacts readily to form disulphide cross-links when
denatured.
Milk proteins are part of the milk solids-not-fat (SNF) content of milk. When skim milk is
concentrated and spray-dried to produce skim milk powder (SMP), proteins comprise
View Online
about 36% of the dry weight of the powder, lactose about 52% and the balance in minerals
and salts. The caseins in SMP are in micellar form and the serum proteins could be in their
native state or denatured, depending on the heat classification of the powder (low to high),
which affects the solubility, heat stability and potential disulphide reactions (with glutens,
for example) of the SMP on reconstitution or application.
The protein content (casein and serum protein) of the SNF can be increased and the
lactose and mineral content decreased by the application of ultrafiltration (UF), a
membrane fractionation process based on hydrated molecular size. The retentate
ingredients are referred to as milk protein concentrates (MPCs), which are available at
protein levels from 42% protein up to 85% protein (dry basis), the latter requiring
intermediate dilution with water in a process referred to as diafiltration. The loss of serum
salts and the use of water modifies the micellar ionic equilibria, which can affect the
aggregation levels of the micelles and hence the solubility and functionality of the proteins.
Milk protein isolates with >90% protein are also available from diafiltration processes.
Caseins are readily precipitated from skim milk by the addition of acid, to pH 4.6,
which is a means of isolating them from all of the other soluble components of SNF.
Centrifugation can then produce acid casein. This insoluble acid casein can be re-
neutralized with alkali, such as sodium hydroxide or calcium hydroxide, to produce
sodium caseinate or calcium caseinate, respectively. Neutralization of pH increases protein
charge and disaggregates the insoluble acid casein. In the case of sodium caseinate, this
results in a soluble and translucent or transparent ingredient with good functionality.
Calcium caseinate results in a white colloidal dispersion, since divalent calcium can
crosslink casein polymers to form small aggregates stable to sedimentation6.
12/04/2014 12:04:43.
Whey is a by-product of the cheesemaking process, the liquid that is drained from curd
after casein micelle aggregation in milk by rennet. The protein content of whey is
comprised of the soluble serum proteins and the caseinomacropeptide of N-casein. Whey
can be concentrated and dried to produce whey powder with about 8-10% protein,
approximately 75% lactose and a high level of soluble milk salts. Due to the low level of
protein, the food applications of whey powder are limited.
The ultrafiltration of whey, followed by evaporation and optionally spray-drying to
powder form, produces whey protein concentrates (WPCs) of up to 35% protein (dry basis)
and the further application of diafiltration produces WPCs up to 80% protein. WPC
powders are generally highly soluble, depending on the various processes applied10.
Protein denaturation is critical to functionality; generally D-lactalbumin and E-
lactoglobulin will not be denatured but the minor proteins may be to varying extents.
View Online
Whey protein isolates, typically >90% protein on a dry basis, can be produced by
ion exchange processing in either a stirred or continuous-flow resin-bed reactor. Acidified,
positively-charged proteins react with the negatively-charged porous resin and
deproteinated whey is removed. Proteins can then be desorbed and eluted from the resin by
neutralization. Selective membrane processing and selective resins can also be used to
isolate a pure stream of D-lactalbumin for specific applications. The resulting WPI is E-
lactoglobulin-enriched but still contains the other protein fractions10.
The permeate from phosphocasein production by microfiltration and diafiltration of
skim milk may be the more desirable end-product of this process, as it contains native
serum proteins that have not been subjected to the various processing steps involved in
cheesemaking. These ingredients have been shown to enhance functionality compared to
their whey-based counterparts7-9.
Proteins that can readily migrate to and stabilize oil or air interfaces, due to their
amphiphilic nature, demonstrate excellent emulsification and foaming properties.
Casein proteins can form stable emulsion droplet interfaces and the interfacial layer
can be manipulated for various applications by the extent of aggregation of the protein.
Sodium caseinate, the least aggregated protein, will form thin, coherent and stable
12/04/2014 12:04:43.
emulsion interfaces with uniform coverage and the lowest adsorbed protein levels (~2
mg/m2 of droplet interface) of all of the casein ingredients, leading to the formation of
large surface area/small droplet diameter during emulsification/homogenization. As the
extent of aggregation of protein increases, from calcium caseinate to micellar casein, the
adsorbed protein content will increase (up to >10 mg/m2 for whole milk proteins) and
steric stabilization will increase but the coverage area decreases and proteins become more
susceptible to displacement by surfactants. Native casein micelles will tend to rearrange at
emulsion interfaces to expose the hydrophobic core to the lipid droplet. Such interfaces can
certainly be stable, as demonstrated by homogenized milk. Interfacial areas not coated by
micelles are likely coated by thin layers of serum proteins or dissociated caseins6,11.
Sodium caseinate will also produce high levels of air incorporation in foams and
the resulting foams will be quite stable. Aggregated proteins like calcium caseinate or
micellar caseins are less able to form and stabilize foams and the foams tend to collapse
more quickly. The presence of any fat in the protein preparation will greatly hinder
foamability6,11.
Whey proteins can act as effective emulsifiers at concentrations as low as 0.5% and
form thin, stable interfacial films at 2-3 mg/m2, compared to aggregated proteins10,11. This
renders them nearly as good as sodium caseinates for applications such as sauces or cream
soups. In whipped emulsions such as ice cream and whipped toppings, however, emulsions
must be stable prior to whipping but undergo destabilization through partial coalescence
during whipping to establish the final aerated and emulsified structure. The use of heat-
aggregated whey proteins has been investigated for such applications, to develop whey
protein-based ingredients that behave more as micellar caseins during interfacial
adsorption at the time of homogenization, so they do not produce emulsions that are as
stable as native whey proteins.
Whey proteins form foams with high air content at 3-4% protein; however good
foam stability is only achieved at higher protein levels, up to 15% protein. Foaming of a
View Online
globular protein followed by heating can produce a heat-set stable foam, as the proteins
first adsorb during whipping and then interact through denaturation by heat. Egg white
proteins are superior in this performance, making excellent meringues. It has been difficult,
however, to use whey proteins as replacements for egg white proteins in meringues and
angel food cakes, as they lead to relatively more bubble disproportionation and loss of
stability during heating10,11.
The interactions between proteins and water and their self-associations in aqueous
solutions lead to viscosity and gelation, and in these roles their behaviour is similar to
other hydrocolloids such as polysaccharides.
Heat stability of milk protein fractions is a very important consideration in their
functionality. Concentrated milk proteins generally have lower heat stability than fresh
milk, with a maximum stability at pH 6.7 but a minimum heat stability at pH 6.9. The
integrity of the casein micelle is critical, at elevated pH N-casein dissociation occurs and
the micelle is more susceptible to calcium-induced precipitation. Heat treatment of milk
prior to concentration improves heat stability by the reaction of E-lactoglobulin with N-
casein, which tends to stabilize the micelle surface. The equilibrium of soluble:colloidal
minerals is also important in trying to optimize heat stability for specific applications11.
Micellar casein solutions, for example concentrated milks or MPC solutions, tend
to display fairly low viscosity due to the non-interacting nature of the aggregated proteins.
Sodium caseinate solutions, however, can produce high viscosity under certain conditions.
12/04/2014 12:04:43.
Interactions between milk proteins and polysaccharides can either lead to polymer
interactions or to polymer incompatibility and phase separation. An example of the former
would be conjugation of whey proteins with polysaccharides through dry-heating and
Maillard reaction. An example of the latter would be the phase separation that occurs due
to depletion flocculation between polysaccharides and casein micelles. This can often lead
to wheying off the formation of a clear serum layer or syneresis when polysaccharide
stabilizers or thickeners are added to dairy-based beverages or puddings.
A unique hydrocolloid interaction is the so-called milk reactivity between casein
micelles, particularly the N-casein hairy layer, and N-carrageenan. This functionality of N-
carrageenan is sufficient to inhibit protein-polysaccharide phase separation at the
macroscopic scale (wheying off) although not necessarily at the microscopic scale. It has
been shown that N-carrageenan stabilizes polysaccharide-containing dairy liquids, ice
cream mix or chocolate milk, for example, by interacting with the surface of the micelle
and self-associations, leading to the formation of a weak gel network sufficient to stabilize
the mircoscopically-separated phase domains from further coalescence and separation; in
essence the resulting structure is a stabilized water-in-water type emulsion. Micellar
casein-N-carrageenan interactions and N-carrageenan-N-carrageenan interactions are both
required for sufficient stabilization, making N-carrageenan singularly unique and a
requirement for most polysaccharide stabilizer blends for dairy applications12.
Another unique polysaccharide application for dairy applications is the use of high
12/04/2014 12:04:43.
methoxyl (HM)-pectin for stabilization of acidified milk drinks such as yogurt beverages.
As the pH decreases, the micelles lose their net negative charge, hence stability, and
protein aggregates in the acidified beverages can precipitate and leave a chaulky texture in
the beverage. It has been shown that HM-pectin adsorbs to the surface of the micelle at low
pH, after the negative charges on the micelle have been reduced, and the negatively
charged pectin molecules re-establish the electrostatic and steric repulsion lost by the
micelles. This maintains the casein micelles and aggregates suspended in the acidified milk
drinks13.
Milk protein ingredients are used in applications to provide nutrient content, to replace
other ingredients, to provide specific functional properties and to make novel food
products. In all applications, the flavour quality of the ingredient is paramount, and most
milk protein ingredients should provide a neutral or bland flavour and not bind or cause
loss of other flavour components present3.
Milk ingredients are used in the manufacture of recombined milk products, which
offer beverage milks and other dairy products to many populations of the world with
insufficient indigenous milk supply. Although SMP, anhydrous milkfat and water have
been the traditional and primary ingredients, novel milk protein ingredients have been
shown to improve functional properties such as heat stability. Ice cream mix is another
dairy product in which the SNF component, traditionally supplied by SMP, is now being
supplied by blended milk protein-based ingredients, some with lower protein content but
higher functionality than SMP. In ice cream mix and in dairy-based emulsions destined for
whipping, such as whipped toppings, proteins play the dual role of emulsification, which
View Online
cannot be too stable as discussed above or the emulsion will not partially-coalesce to form
good fat structure, and foaming, to form small and stable air bubbles. One difficulty with
trying to supply a good foaming protein in the formulation is the sequential nature of the
process, where homogenization of the fat occurs before foaming. An amphiphilic protein
that makes a good foam stabilizer will also act as a good emulsifier, and therefore likely to
adsorb to the fat interface and not be available for foaming. This makes control of protein
adsorption to the fat interface critical to the development of good foam structure. Hence
the use of small molecule surfactants, which primarily promote fat structuring, also
indirectly promote good air bubble formation and stability, both through the primary
adsorption of available proteins to the air interface and through the secondary adsorption of
partially-coalesced fat to the air interface14.
A specific application of whole milk proteins is in the manufacture of milk
chocolate. In this instance, whole milk powder is preferred (~25% in milk chocolate), and
more specifically roller-dried powder with a high level of free fat, but skim milk powder
and whey protein ingredients have also been used as have modified spray-dried whole milk
powders to enhance fat aggregation. Chocolate is fat-continuous with no water, so in this
application the role of casein micelles seems to be to interact with sugar and modify the
chocolate fat rheology6,11.
One of the principal applications of caseinate ingredients is to create non-dairy
products such as coffee whiteners, where good emulsion stability and heat stability is
required, and whipped toppings, where good foaming properties are required but in such
products it is important than fat emulsion droplets not be too stable or they will not form
optimal structures through partial coalescence. Another unique application of sodium
12/04/2014 12:04:43.
caseinate is in cream liqueurs, to offer high emulsion stability and block fat creaming and
plug formation in the neck of the bottle during ambient temperature shelf-life of several
months6.
Whey protein ingredients can provide specific nutrients in products like infant
formulae, to balance the amino acid composition of whole milk proteins to more closely
simulate human milk protein: a balance of 60% whey protein and 40% whole milk protein
is often used. Isolated sources of D-lactalbumin are also utilized to create more closely
balanced infant formulae, since E-lactoglobulin does not exist in human milk and is often
the source of milk allergies. Once hydrolysed, however, the allergenic structures are
broken down, suggesting good opportunity for commercial production of whey protein
hydrolysates. Whey proteins are utilized in sports nutrition applications because of their
high level of branched-chain amino acids, which are preferentially metabolized by muscle
rather than in the liver. Whey proteins are often used in enteral nutrition products to
contribute relatively high levels of essential amino acids at low protein loads10.
In bakery applications, whey proteins can interact with gluten proteins though
disulphide bonds to modify gluten rheology. High-heat milk powders are preferred, to
induce whey protein denaturation and expose the sulfhydryl groups10.
6 CONCLUSIONS
The dairy industry has many years of experience at isolation of ingredients from milk that
are enriched in protein, and the research literature on milk protein functionality and
applications is extensive. Fractionation continues to increase the range of specific protein
products available, driven by improvements in functionality for unique applications. Novel
processing methods such as high hydrostatic pressure processing of milk or milk
ingredients/products, enzymatic modifications of specific milk proteins, for example the
View Online
References
1 P.L.H. McSweeney and P.F. Fox, eds. 2013. Advanced Dairy Chemistry. Vol. 1A:
Proteins Basic Aspects, 4e. Springer, New York.
2 P.L.H. McSweeney and S. OMahony, eds. 2013. Advanced Dairy Chemistry. Vol.
1B: Proteins Applied Aspects, 4e. Springer, New York.
12/04/2014 12:04:43.
3 P. Walstra, J.T.M. Wouters and T.J. Guerts. 2006. Dairy Science and Technology, 2e.
CRC/Taylor & Francis, Boca Raton, FL.
4 D.G. Dalgleish. 2011. On the structural models of bovine casein micellesreview and
possible improvements. Soft Matter, 7, 22652272.
5 D.G. Dalgleish and M. Corredig. 2012. The structure of the casein micelle of milk and
its changes during processing. Annu. Rev. Food Sci. Technol., 3, 44967.
6 B.T. OKennedy. 2011. Caseins. In Handook of Food Proteins. G.O. Phillips and P.A.
Williams, eds. Woodhead, Cambridge, UK.
7 B.K. Nelson and D. M. Barbano. 2005. A microfiltration process to maximize removal
of serum proteins from skim milk before cheese making. J. Dairy Sci., 88, 1891-1900.
8 E. Hurt and D.M. Barbano. 2010. Processing factors that influence casein and serum
protein separation by microfiltration. J. Dairy Sci., 93, 4928-4941.
9 M.C. Adams and D.M. Barbano. 2013. Serum protein removal from skim milk with a
3-stage, 3x ceramic isoflux membrane process at 50C. J. Dairy Sci., 96, 2020-2034.
10 M. Boland. 2011. Whey Proteins. In Handook of Food Proteins. G.O. Phillips and
P.A. Williams, eds. Woodhead, Cambridge, UK.
11 M.A. Augustin and P. Udabage. 2007. Influence of processing on functionality of
milk and dairy proteins. Adv. Food Nutr. Res., 53, 1-38.
12 P. Spagnuolo, D.G. Dalgleish, H.D. Goff and E.R. Morris. 2005. Kappa-carrageenan
interactions in systems containing casein micelles and polysaccharide stabilizers.
Food Hydrocoll., 19, 371-377.
13 R.H. Tromp, C.G. de Kruif, M. van Eijk and C. Rolin. 2004. On the mechanism of
stabilisation of acidified milk drinks by pectin. Food Hydrocoll., 18, 565-572.
14 H.D.Goff. 2013. Milk Proteins in Ice Cream. In Advanced Dairy Chemistry. Vol. 1B:
Proteins Applied Aspects, 4e. P.L.H. McSweeney and S. OMahony, eds. Springer,
New York.
GELATINS PHYSICOCHEMICAL PROPERTIES, SOURCE DEPENDENCE AND
APPLICATIONS
1 INTRODUCTION
With an estimated annual world-wide production of more than 300 000 metric tonnes,
gelatin is one of the most abundant commercial biopolymers, and with several hundred
12/04/2014 12:04:45.
different applications, ranging all the way from technical to pharmaceutical, it is also one
of the most versatile. Gelatin, originating from Latin gelare meaning to gel / solidify, has
been known from ancient times and has undergone intensive research in the recent 60 years
as described in comprehensive monographs by Veis1, Ward and Courts2 and Schrieber and
Gareis3. Production of gelatin is based on a partial hydrolysis of the mother structural
molecule collagen, mainly originating from mammalian sources. But as collagen is one of
Natures evolutionary success stories it is found all over the animal kingdom and hence
several possible sources exists (see below). Gelatins from mammalian (homoeothermic)
sources have basically identical amino acid composition as well as physical properties,
whereas the physico-chemical properties of gelatins (and collagens) from poikilothermic
animals may vary considerably.
2.1 Sources
Hide, skin and bone from mammalian (bovine and porcine) sources are by far the most
used and also preferred raw materials, due to a steady supply and well-established quality
control. Today, gelatins from a variety of fish species as well as a small quantity of avian
gelatins are also manufactured. However, fish and poultry sources only account for 2 3%
of the total annual gelatin production. This is partly due to a lack of stability with respect to
availability of good quality supply, but another major issue is that gelatins from cold water
fish species have sub-optimal properties compared to mammalian gelatins; both with
respect to gelling and melting temperatures as well as the mechanical properties (gel
strength) of the final gels. Gelatins from warm water fish species, on the other hand, have
almost mammalian-like properties and can directly replace mammalian gelatins in many
applications.
View Online
2.2. Manufacturing
All biological materials intended for gelatin production are thoroughly rinsed to remove
impurities and de-fatted if so required. Following this cleaning treatment, the raw material
is pre-treated with either acid (typically pork or fish raw material) or alkali (typically
bovine raw material) prior to gelatin extraction. These pre-treatments are performed to
break inter-molecular covalent bonds within and between the triple helical collagen
although intra-molecular bonds may also be broken to a certain extent. Acid pre-treatment
implies that the raw material is immersed in cold and dilute mineral acid for typically 18
24 hours, and results in Type A gelatin. Alkaline pre-treatment typically implies the use of
saturated lime water for a period of months
From the pre-treated raw material gelatin is extracted through a series (typically 3 to
5) of hot water treatments with temperatures increasing successively in the range from 50
to 100 C. As a result of both the pre-treatment and the extraction procedure the
manufactured gelatin will inevitably be a mixture of polypeptide chains of different
molecular weights and hence differ from the more typical monodisperse, globular proteins.
One triple helical collagen molecule consists of three -chains with a molecular weight of
around 100 kDa, but a gelatin product will contain both higher molecular weights (- and
-chains; i.e. two and three linked -chains, respectively) as well as sub- peptides
resulting from intramolecular hydrolysis. Typically, the fraction of sub- increases with
increasing extraction temperature and time.
12/04/2014 12:04:45.
3. PHYSICO-CHEMICAL PROPERTIES
Each of the three -chains making up the collagen monomer (tropocollagen) has a general
amino acid sequence Gly-X-Y, where X usually is proline and Y is hydroxyproline. Hence,
glycine accounts for around 30% of all the amino acid (AA) residues. Type A gelatin has
an AA sequence virtually identical to the parent collagen, and an isoelectric point (pI) in
the range 8 9. Type B gelatin, on the other hand and due to the alkaline pre-treatment,
has a lower pI of around 5. The alkaline pre-treatment leads to a deamination of the acid
amides glutamine and asparagine to their respective acids (glutamic and aspartic acids).
The -chains can be looked upon as being composed of both polar and non-polar regions
where the Gly-X-Y represents the latter. Due to this particular heterogeneity with respect
to polarity, gelatin is a multifunctional hydrocolloid with surface activive properties (see
later).
The total fraction of proline plus hydroxyproline vary quite markedly between gelatins
from different sources. For mammalian gelatins (homoeothermic animals), this content is
stable between species with approximately 220 residues per 1000 AA. For cold water fish
gelatins it may be as low as 150, whereas for warm water fish gelatins it is around 200. 4
These variations have significant impact on the physical properties.
It is generally acknowledged that the thermal stability of collagen as well as the physical
properties of gelatin is determined by the contents of the imino acids proline and
hydroxyproline5-9 as they are pivotal to the stabilisation of the triple helical structure. The
content of these imino acids hence determines the helix-to-coil transition temperature, a
View Online
most critical gelatin parameter, and is linked to the breaking of inter- and intramolecular
hydrogen bonds. Above this transition temperature, the gelatin will exist as random coils in
solution. As gelatins are rather flexible coils above the transition temperature with intrinsic
viscosities ([]) typically between 40 and 50 ml/g,7 rather concentrated solutions (e.g.
25%) of gelatins can easily be manufactured.
The Bloom value is an industrial standard that essentially measures the rigidity of a
gelatin gel under standard conditions, and the given Bloom value represents the weight (as
force in grams) necessary to compress the surface of a gelatin gel of a defined geometrical
shape 4 mm using a flat bottomed cylindrical plunger. As such, the Bloom value does not
represent any fundamental rheological entity, but remarkably good correlations have been
found between Bloom values and the dynamic storage modulus (G) when the gels have
experienced the same history.10
As can be seen from figure 1 there are huge effects on the development of the
dynamic storage modulus when gelatins from different sources, and hence also different
content of imino acids, are observed at standard conditions for Bloom strength incubation
(6.67 wt.% gelatin, 10 C). The CWF gelatin in figure 1 is a particularly high molecular
weight gelatin isolated under mild conditions (extracted at room temperature).
Commercially available gelatins from such sources do normally not show any transition at
all under these conditions due to a remarkably low setting temperature (4 7 C ) because
of a low imino acid content as well as a suboptimal molecular weight. The WWF gelatin,
on the other hand, exhibits mechanical properties superior also to the presented
mammalian gelatins. Also this sample was optimised with respect to molecular weight
during pre-treatment and extraction.
12/04/2014 12:04:45.
10000
9000 WWFG
Bloom 275
8000
7000
6000 Bloom 205
CWFG
G' (Pa)
5000
4000
3000
2000
1000
0
0 6 12 18 24
Time (h)
Figure 1: Development of the dynamic storage modulus (G) for 4 different gelatins during
18 hours incubation at 10 C. CWFG = cold water fish gelatin isolated from North Sea
saithe at 20 C. WWFG = warm water fish gelatin isolated from Vietnamese Pangasius at
35 C . Bloom 205 and 275 represent two mammalian gelatins (bovine, type B).
View Online
Even though alternative raw material to mammalian gelatin exist, which would not
have any cultural limitations, the major challenges with e.g. warm water fish gelatins is
control and quality of the raw material as it is easily microbially contaminated and there is
no tradition in considering fish skins as high quality products. Furthermore, the availability
of such products inevitably will depend upon variable factors like fisheries regulations
(living stock considerations) and by-product availability. It therefore seems likely that such
gelatins will be supplements, rather than competitors, to their mammalian counterparts.
In order to improve the physical properties of CWFGs several possible approaches
have been suggested. A possible method is the use of transglutaminases to introduce
intermolecular covalent linkages to improve connectivity. The main challenge with this
method is to be able to control the amount of cross-links because the typical melt-in-
mouth texture of gelatin is easily lost as the amount of covalent cross-links is increased. In
addition, if the degree of cross-linking is too low, an acceptable sol-gel transition
temperature of the modified gel may not be achieved or the gel strength would be weak.11
Hence, the enzyme activity needs to be carefully controlled in order to obtain CWFG gels
with acceptable rheological properties. An alternative approach for enhancing the gel
firmness is the use of chemical cross-linkers, such as glutaraldhyde, genipin and
carbodiimides.12 The chemically cross-linked gels would however have the same
disadvantages as the enzymatically treated gels, in which the crosslinking would interfere
with the thermoreversibility and fracture stress of the gelatin gel and would hence not give
the same characteristic texture during e.g. mastication. It should also be noted that
chemical crosslinkers are often not considered to be very biocompatible.
The use of mixed biopolymer systems have been suggested as a possible means for
12/04/2014 12:04:45.
obtaining gels with improved melting and gelling temperature and gel strength. The use of
-carrageenan and CWFG to prepare mixed biopolymer gels have been reported.13 In this
study, a high level of complexity was observed with both polymers undergoing dis-
order/order transitions and hence also temperature dependent and varying phase behaviour.
An overall increase in elastic properties was observed at low temperatures. Mixed
biopolymer films of WWF gelatin and -carrageenan or gellan have also been
investigated.14 An increased melting temperature was obtained for both systems, with
gellan being more effective. Besides chemically/enzymatic crosslinking or mixed
biopolymer gels, the addition of both co-solvents15 and salts9, 15 have been reported to give
fish gelatin gels with improved melting temperature. The usage of such coenhancers is
somewhat limited for use within the food industry due the high concentration needed of the
salt or co-solvent.16
4. APPLICATIONS
As previously mentioned, gelatin may also be used to form filled gel emulsions, in
which oil droplets are dispersed throughout a gelatin network. Different types of food and
confectionary can be described as emulsified gelatin gels, in addition such gels have also
been suggested as a possible delivery vehicle for lipid-based nutraceuticals and
pharmaceuticals.24, 25 The amphipathic property of gelatin enables it to adsorb at the oil-
water interphase causing connectivity between the dispersed droplets and the continuous
gel network, also known as active fillers. It has been suggested that if the modulus of the
droplets is higher than the bulk matrix, such matrix-filler interaction may promote a
reinforcement of the gelatin network, leading to an increase in gel firmness.26 An early
study suggested that a critical concentration of gelatin is needed to achieve an increase in
storage modulus, since the peptides adsorbing at the oil-water interphase will not
participate directly in bulk gelation.27 This suggests that both the droplet size and the
volume fraction of the dispersed phase may influence the mechanical properties of filled
gel emulsions.
The interaction between lipid droplets and the gelatin network is hence of
considerable importance. By adding an alternative emulsifying agent, gelatin may be
displaced from the droplet interphase. This may lead to a reduced interaction between the
droplets and gelatin network, causing the dispersed phase to become an inactive filler. For
such filled gel emulsions a decrease in gel firmness would be observed. The difference
between inactive/active fillers is clearly illustrated by studies performed by Hattrem,
Molnes and Draget,28 in which an increase in storage modulus (measured at 20 C after 15
minutes of curing) was measured for a filled gel emulsion (40 wt.% corn oil) prepared with
a pig skin gelatin (260 Bloom) compared to a gelatin gel. As expected, the introduction of
12/04/2014 12:04:45.
the water soluble surfactant, polysorbate 80, reduced the storage modulus of the filled gel
emulsion dramatically by replacing the gelatin from the droplet interphase leading to the
lipid droplets becoming an inactive filler. These data are presented in figure 2.
45000
40000
35000
30000
G' (Pa)
25000
20000
15000
10000
5000
0
0 % oil 40 % oil
MQ Tween80
Figure 2: Measured G (Pa) after 15 minutes of curing at 20 C for gelatin gels and filled
gel emulsions (40 wt.% corn oil) prepared using 260 g Bloom gelatin type A (25 wt.% of
the continuous water phase) in the presence or absence of 0.5 wt.% polysorbate 80 (tween
80) determined by small strain oscillatory measurements during a cooling-heating process
from 60 C 20 C (15 of minutes of curing) 60 C with a temperature gradient of 2
C/min.
View Online
7000 70
6000 60
5000 50
Temp (C)
4000 40
G' (Pa)
3000 30
2000 20
1000 10
0 0
0 10 20 30 40 50 60 70
Time (min)
G'(Pa)-Gelatin gel (0 wt.% oil) G'- Gelatin gel (40 wt.% Corn oil) Temp
Figure 3: Measured G (Pa) for gelatin gels and filled gel emulsion (40 wt.% corn oil)
prepared using high molecular weight cold water fish gelatin (25 wt.% of the continuous
water phase) determined by small strain oscillatory measurements during a cooling-
heating process from 60 C 10 C (15 minutes of curing) 60 C with a temperature
gradient of 2 C/min
View Online
5 CONCLUSION
The lower fraction of imino acids in collagen extracted from poikilothermic animals living
in cold habitats, compared to those living in warmer habitats or homoeothermic animals,
gives gelatins with sub-optimal gelling properties. This is reflected by the lower gelling
temperature of cold water fish gelatins (CWFGs) compared to gelatin extracted from
porcine or warm water fish sources. Different approaches have been used for optimising
the rheological properties of CWFG gels, such as careful extraction of gelatin from the
collagen source to avoid intramolecular hydrolysis, chemical or enzymatic cross-linking,
mixed biopolymer gels or the addition of salts or co-solutes. Gelatins have a wide area of
applications, mainly connected to its gelling, amphipathic or ampholytic properties. Fish
gelatins may be used for applications connected to both of the latter properties as these are
not connected to the imino acid content. Provided a connectivity between the gelatin
matrix and the dispersed phase a reinforcement of the gelatin gels may be achieved. Hence,
a CWFG gel with improved gel firmness can be manufactured by introducing dispersed
droplets.
References
1. A. Veis, The macromolecular chemistry of gelatin, Academic Press, New York,, 1964.
2. A. G. Ward and A. Courts, The Science and technology of gelatin, Academic Press,
London ; New York, 1977.
3. R. Schrieber and H. Gareis, Gelatine handbook : theory and industrial practice,
12/04/2014 12:04:45.
22. S. K. Kim, Y. T. Kim, H. G. Byun, K. S. Nam, D. S. Joo and F. Shahidi, J Agr Food
Chem, 2001, 49, 1984-1989.
23. B. Gimenez, A. Aleman, P. Montero and M. C. Gomez-Guillen, Food Chem, 2009,
114, 976-983.
24. I. J. Haug, L. B. Sagmo, D. Zeiss, I. C. Olsen, K. I. Draget and T. Seternes, Eur J
Lipid Sci Tech, 2011, 113, 137-145.
25. A. C. K. Sato, K. E. F. P. Moraes and R. L. Cunha, Food Hydrocolloid, In Press.
26. T. Vanvliet, Colloid Polym Sci, 1988, 266, 518-524.
27. E. Dickinson, G. Stainsby and L. Wilson, Colloid Polym Sci, 1985, 263, 933-934.
28. M. N. Hattrem, S. Molnes and K. I. Draget, In Prep.
29. F. Madsen, Roy Soc Ch, 2000, 411-420.
30. J. A. Omotosho, Int J Pharmaceut, 1990, 62, 81-84.
31. T. Yoshioka, K. Ikeuchi, M. Hashida, S. Muranishi and H. Sezaki, Chem Pharm Bull,
1982, 30, 1408-1415.
12/04/2014 12:04:45.
PROPERTIES AND APPLICATIONS OF SOY PROTEINS
1. INTRODUCTION
Soybeans have been cultivated for more than 3000 years in China and other Asian
countries, such as Japan and Korea. Some trials to cultivate soybeans have been known in
France and England since the 18th century, but have not been developed further. Since
12/04/2014 12:04:47.
1930, USDA has developed the cultivation and now USA has the largest production in the
world: USA, 7107t; Brazil, 5.8107t; Argentine, 5.8107t; China, 1.7107t; India,
1.0107t 1. Soybeans have been an important protein source in Asian countries and have
been utilised in various forms such as tofu (soybean curd), miso (fermented soybean paste),
natto (fermented soybeans covered with mucilagenous substance), aburage (fried sheet of
tofu) etc 2. Recipe books on more than 100 different tofu dishes were published in the Edo
era (18th century) in Japan. In addition to these traditional foods, an increased amunt of
soybean milk is now consumed in Japan and in China due to its expected health benefit.
Fibrous texture was also introduced in tofu-like foods, making it resemble meat-like foods.
Chen, Yamaguchi and Ono 3 recently shed new light upon the formation of yuba, a film-
like soybean food made from heated soymilk that contains oil bodies, particulate protein,
soluble protein, and carbohydrate.
The advantages of soybean proteins are :1) provides a good balance in amino acid
composition, since all the essential amino acids are contained, 2) contains physiologically
beneficial components which are shown to lower the cholesterol, and reduce the risk of
hyperlipidemia and cardiovascular diseases, 3) has excellent processing ability such as
gelling, emulsifying ability and water- and oil- holding capacity.
Soybeans should be heated before use in order to 1) deactivate physiological harmful
substances, such as trypsin inhibitor, and hemaglutinin, 2) induce the denaturation of
soybean protein, 3) soften the tissue of soybean, 4) remove or reduce the raw soybean odor,
5) to sterilize 4.
In addition to protein and oil, physiologically beneficial effects of daizein, isoflavone in
soybeans have been attracting much attention 1. Soluble soybean polysaccharides extracted
from residue (okara) in tofu-curd production have been shown to be a good emulsifier and
have been widely used in the food industry 1.
View Online
Soybean contains approximately 40% protein and 20% oil on an average dry matter base.
By removing oil at lower temperatures, soy protein isolate (SPI) is obtained, and is widely
used in the food industry. Whole aqueous extractable soybean proteins can be separated
into storage globulin and whey fractions by acidification to pH 4.5-4.8. The acid
precipitable fraction includes the major soybean storage proteins, and which is the main
material considered in the present chapter. The remaining part consists of the minor
globulin -conglycinin, and relatively large amounts of contaminating proteins, including
whey proteins which make up 9-15.3% of soybean protein 5. Whey proteins are composed
of lipoxygenase (LOX, 102 kDa), -amylase (61.7 kDa), lectin (33 kDa), and Kunitz
trypsin inhibitors (KTI, 20 kDa) 6. The proportion represented by these whey proteins in
the acid precipitated globulins is unknown 7.
SPI is a mixture of various proteins, and the main ingredients are classified into four
protein categories according to their sedimentation coefficients 2S, 7S, 11S and 15S which
sediment at different gravitational forces when the solution is subjected to a centrifugal
field. Among these four proteins, 7S (-conglycinin) and 11S (glycinin) represent more
than 80%, and the ratio 7S/11S has been reported to be about 0.5 to 1.3 depending on
varieties 8.
7S globulin consists of three subunits (ca 67 kDa), (ca 71 kDa) and (ca 50 kDa). 11S
globulin is a hexamer, and is made up of five different subunits, each of which consists of
an acidic subunit A (acidic pI) with a molecular mass about 35kDa and a basic subunit B
(basic pI) of molecular mass about 20kDa, linked by a disulfide bond. AB subunits are
12/04/2014 12:04:47.
believed to associate into two hexagonal rings forming a hollow cylinder by electrostatic
and hydrogen bondings 9. Glycinin (11S) was found to dissociate into 2S, 3S or 7S forms
in various pH and ionic strengths.
Amino acid compositions of -conglycinin and glycinin have been analysed, but
crystallization was difficult and the three dimensional structure is not well established 10 in
spite of many efforts. The crystal structure of 7S and 11S have been recently studied by X-
ray diffraction 11, and the previously proposed picture of 11S was reconfirmed and refined.
They proposed that the movement of a mobile disordered region to the side of the trimer,
and the dissociation of the hexamer into trimers may be susceptible to proteinases 11b.
Native glycinin is known to have a compact structure stabilized by disulfide bonds and
thus its emulsifying and foaming ability is lower than that of -conglycinin which lacks
disulfide bonds.
Ren et al 12 analyzed the aggregation mode of polypeptides in protein particles of soy milk
by using ultracentrifugation, gel filtration, and sodium dodecyl sulfate polyacrylamide gel
electrophoresis. They proposed the interaction mechanism of polypeptides in heat-induced
protein particles of soy milk: The proteins in soy milk dissociated, rearranged, and
aggregated to form protein particles when heated. The protein particles of >40 nm in
diameter dissociated into protein aggregates with various molecular masses, which were
dissociated into monomeric subunits of 7S and 11S protein after treatment by the mixture
of 6 M urea and 0.5% SDS. The aggregates were primarily composed of the disulfide-
linked basic and acidic polypeptides of 11S, besides a very small amount of and
subunits of 7S. These aggregates and a part of monomeric subunits of 7S and 11S, as
structural units, interact with each other to form protein particles primarily via noncovalent
interactions, especially hydrophobic interactions and hydrogen bonding. The disulfide-
linked basic polypeptides of the aggregates should be located inside the protein particles,
View Online
whereas the acidic, and subunits should be located outside them for their high
hydrophilicity.
Recently, the classification of soybean globulin was contested, and the existence of a
lipophilic protein (LP) in addition to 7S and 11S was proposed 13. The protein content of
LP was reported to be 76% which is lower than the 87% of 7S and 93% of 11S. However,
LP showed a higher content of lipid 11.7% than 0.8% of 7S and 3.3% of 11S. Samoto et
al 13 reported that the LP yield decreased, and simultaneously the yield of residue increased
although the yield of 7S and 11S was not changed with increasing temperature. They
attributed this change to the acceleration of aggregation of LP by heating because of the
hydrophobic properties of LP.
Wu et al 14 reported a pilot plant fractionation to produce kilogram quantities of 11S, 7S
and an intermediate mixture using ultrafiltration rather than acid precipitation. They also
reported the effects of reducing agents and salts concentration on the fractionation, yield
and purity of soybean storage protein fractions 15.
Glycinin was recently isolated from soybeans using a monoclonal antibody with a yield of
16.8% and a purity of 93.8% based on immunoaffinity chromatography, which were
significantly higher than those produced using other traditional procedures 16.
3. FUNCTIONALITY
3.1 Solubility
12/04/2014 12:04:47.
During soymilk processing, the soy proteins are prone to denature and tend to interact with
each other when heated. Many research groups have investigated the aggregation of soy
proteins using pure proteins. The effect of 7S on the thermal aggregation of 11S has been
examined. The addition of isolated 7S has been reported to prevent the thermal aggregation
of both 11S and the isolated basic subunits of 11S 17. The preferential association of the
basic subunits of 11S and the subunits of 7S through an electrostatic force when 11S and
7S are heated together has been reported 18. The N-linked carbohydrate moieties of the
and subunits of 7S could prevent the heat-induced associations of 7S 19.
Isolated glycinin aggregates when heated at 80C 20. The ultracentrifugal pattern showed
that heating of acid-precipitated soy protein at 80C resulted in the disappearance of
glycinin and the concurrent appearance of soluble aggregates and protein components of 2-
4 S. 17 found a similar function for egg albumin and bovine serum. They attributed the
prevention of aggregation by -conglycinin to the formation of a soluble complex between
the subunits of -conglycinin and the basic subunits of glycinin via electrostatic interaction.
Guo et al 21 studied the thermal aggregation kinetics of -conglycinin and glycinin at pH
7.0 using size exclusion chromatography with low-angle light scattering (SEC-LALS) to
reveal the assembly process of the soy protein unfolding polypeptide chains during heating
and clarify the role of -conglycinin in the thermal aggregation of soy protein. They also
characterized the structure of the aggregates by small-angle X-ray scattering (SAXS) and
dynamic light scattering (DLS), to know the conformational difference between soluble
and insoluble aggregates. They found that size of both -conglycinin and glycinin
increased with increasing heating temperature, and the size of the mixture glycinin/-
conglycinin decreased with increasing -conglycinin content. Unlike the -conglycinin
soluble aggregates that possessed limited size and less compact conformation, particles
with a denser core and a less dense outer shell were found in the glycinin insoluble
aggregates. They showed that growth of the size of the insoluble aggregates of glycinin
View Online
NSI =
calcium 25. Although there have been many studies on the solubility of soybean proteins
using fractionated 7S and 11S, it is also necessary to understand the interaction between
different ingredients in soybeans. 26a studied the interaction between whey soybean protein
(WSP) and 7S during the formation of protein particles at elevated temperatures. This
indicated that 7S and WSP interact with each other during heating and form complex
protein particles which have special surface properties that are different from individual 7S
and WSP protein particles. They have also shown that lipoxygenase (LOX), -amylase,
and lectin of WSP and the subunit of 7S tend to form a particulate fraction, while the
Kunitz trypsin inhibitors (KTI), , and subunits tend to form a soluble fraction. To
clarify which proteins increase protein particles in heated WSP with increasing
temperature from 65C to 75C, the changes in soluble protein constituent in heated WSP
after being heated from 60 to 95C were analyzed by SDS-PAGE.
It has been reported that the ionized carboxyls have an absorption band around 1570cm-1,
whereas protonated carboxyls are characterized by a band near 1710cm-1. Nagano et al 48c
View Online
found that the different spectra at acidic pHs with the spectrum at pH 7.0 indicated that the
ionized carboxyls had an absorption band at 1562cm-1 while protonated carboxyls were
characterized by a band at 1718cm-1 respectively. The degree of protonation of the
carboxyl groups of -conglycinin was shown to increase with decreasing pH below pH 6.0.
FTIR spectra for -conglycinin gels showed that a band at 1620cm-1 (associated with
exposed -strands) developed with decreasing pH suggested that -conglycinin undergoes
denaturation with increasing protonation of its carboxyl groups, which leads to an increase
in the amount of exposed -strands. The exposed -strands then intermolecularly bond to
form gel networks. This process is promoted with decreasing pH, and as a result, rigid gels
form at acidic pH values.
Nagano et al 48c showed that denaturation peak temperature of -conglycinin detected in
heating DSC shifted to lower temperatures with decreasing pH. Since the denaturation is
the prerequisite of gel formation for globular proteins, it is consistent with the experimental
finding that the storage modulus increased with decreasing pH. A similar experimental
finding that the storage modulus increased with decreasing pH was shown for 12wt%SPI
in the presence of 0.2MNaCl 49. They also found that denaturation peak temperature of -
conglycinin in SPI detected in heating DSC shifted to lower temperatures with decreasing
pH.
The same rheological measurement showed that sodium chloride retards the gelation
process of -conglycinin. Nagano et al 39b found that the gelation time became longer and
12/04/2014 12:04:47.
the gelation rate became slower for -conglycinin with increasing NaCl concentration by
measuring the complex shear moduli at 80C. This finding is consistent with the
stabilization of protein molecules by sodium chloride to heat denaturation. Wongprecha et
al 50 showed that the DSC endothermic peak accompanying denaturation of soy glycinin
shifted to higher temperatures with increasing concentration of added NaCl. Lakemond et
al 51 also found that the denaturation temperature of soy glycinin shifted to higher
temperature with increasing ionic strength by DSC measurements. Similar effects of salt
on the gelation can be found for 7w/w% -lactoglobulin, an important milk whey protein,
in solutions of sodium chloride and calcium chloride 52. The increase of the storage
modulus on heating at 80C was found to delay with increasing ionic strength. It was also
shown that the storage modulus of -lactoglobulin continued to increase after the
temperature was lowered from 80C to 25C. Nagano et al 39b also showed that the gelation
of -conglycinin was retarded with increasing added guanidine hydrochloride (GuHCl) and
that no gelation occurred in the presence of 4M GuHCl.
ovalbumin (1.56-1.80) and casein (1.93-2.0) tabulated by van der Linden and Sagis. These
authors stated that using a percolation model rather than using a fractal model where C0 is
assumed to be zero is more reasonable to establish a generic description.
In the gel formation of globular proteins, some portion which is not completely denatured
may be incorporated in gel network, and in such a situation the effective concentration
contributing to the elasticity may be lower than the total protein concentration. This should
be taken into account to discuss the concentration dependence of elastic modulus of
globular protein gels.
Most commonly used coagulants to make tofu are glucono-delta-lactone (GDL) and
calcium sulphate (CaSO4). Kohyama et al 55 studied the gelation of SPI using these
coagulants, and found that the gelation by calcium was faster than by GDL and the main
molecular forces are hydrophobic interaction. The value of the saturated value of the
storage modulus for 5% SPI increased with CaSO4 concentration up to 35 mM and then
decreased while the rate constant k of the gelation continued to increase at higher
concentration than 35mM. This accelerating gelation action of the coagulant calcium
sulphate (CaSO4) was similar as observed with GDL 53,56. Higher concentrations of CaSO4
than 35 mM led to larger aggregates and syneresis of the gels; therefore, the apparent value
of the saturated storage modulus decreased. They proposed a gelation mechanism of SPI in
the presence of these coagulants; after the exposure of hydrophobic groups by heat
denaturation soluble aggregates are formed, and then the protons from GDL or calcium
12/04/2014 12:04:47.
ions from CaSO4 shields the electrostatic repulsion of negative charges on the surface of
soluble aggregates and leads to the network formation. Essentially both play a similar role
in spite of kinetic difference, and it is consistent with the microscopic observation which
shows a similar microstructure 57. Liu et al 58 recently found similar results for GDL and
CaSO4 although they found that the network of SPI gels prepared with GDL was more
compact and smoother than those with CaSO4. They attributed the difference to the rapid
dissociation of CaSO4 molecules at low temperature and a shorter gelation time.
Nagano examined the difference of fracture behavior and structure of tofus prepared by
MgCl2 and GDL 59. He found that networks of tofu using soy Fukuyutaka prepared by
GDL is much denser than that prepared by MgCl2. Since the gelation rate is much faster
when magnesium chloride is used than when GDL is used, the processing is easier for
GDL and also with less syneresis. Therefore, he could get tofu enough firm which was
self-standing with 7 times water by using magnesium chloride and 14 times water by using
GDL.
gelation will make tofu more retort-resistant, i.e., suppression of water release, and
hardening of tofu pieces packed with water in a retort-pouch. They found that soy proteins
have been polymerized by TGase.
(6% to 9%) than critical concentration for gelation could form a gel by adding calcium
chloride. Above the critical concentration, the heat denaturation leads to gelation as
described above. They used calcium induced cold-set soybean gels as a delivery tool of a
heat labile nutraceutical compound riboflavin 67. They prepared two different gels of 9w/w
SPI by changing calcium concentration, filamentous (10mM CaCl2) and particulate (20mM
CaCl2), and examined the swelling behaviour of these gels in simulated gastric fluid (SGF)
and simulated intestinal fluid (SIF). They found that gels swelled in SGF while they first
swelled but later they shrunk and then collapsed in SIF. They correlated this swelling
behaviour with release of riboflavin and concluded that calcium induced cold-set soybean
gels could be used as vehicles for entrapping bioactive molecules to be delivered and
absorbed in the intestines.
Speroni et al 68 reported high pressure denatured dispersions of SPI, a -conglycinin
enriched fraction (7SEF) and a glycinin enriched fraction (11SEF) with lower
concentration also formed self-standing cold-set gels by subsequent calcium incorporation,
and suggested the possibility of incorporating heat-labile compounds or probiotics during
the gelation step. 7SEF formed aggregated gels with low water holding capacity whereas
11SEF did not form self-standing gels. SPI formed the better gels: ordered and with high
water holding capacity. It is necessary to take into account the different sensitivity of each
protein to high pressure treatment. Denaturation of each protein by high pressure treatment
is different from that by heat treatment; 11S is 100% denatured after a 400 MPa treatment,
while 7S conserves about a 30% of native structure after a 600 MPa treatment. This is in
line with 65 but contradicts 64.
the smooth texture is lost, therefore, the better processing was investigated by many
research groups 69. With decreasing freezing temperature, the pore size of the network
became smaller as expected because the temperature was lowered faster when tofu was put
at lower temperature thus passing faster the temperature range of maximum ice crystal
formation.
Jambrak et al 70 reported that the solubility of soy proteins is increased after ultrasound
treatment, which was attributed to the unfolding and breaking of peptide bonds by
hydrolysis. Hu et al 71 reported that high intensity ultrasonic pre-treatment (HUS) of SPI
improved the water holding capacity and gel strength of GDL-induced-SPI-gels (GISG).
They showed that HUS pre-treatments reduced particle size, increased surface
hydrophobicity of SPI and formed soluble aggregates, leading to denser and more uniform
GISG, and thus the potentiality in food industry.
SPI is widely used in brine for injected salt soluble meat gel products such as ham and
roast beef to maintain texture and retain moisture. It is generally known that the gel
strength of polymer gels decreases with decreasing molecular mass. However, it was
shown that enzyme hydrolysed 7S globulin could increase the gel strength of salt-soluble
meat protein gel in comparison with non-hydrolysed SPI 23.
Tofu production without producing residues and waste water, which are also rich in dietary
fibre, using whole soybean powder has attracted much attention. CLSM observation of
soybean curd made from powdered whole soybean and normal tofu with approximately the
same protein concentration showed that the latter tofu has a finer and a more homogeneous
network than the former curd suggesting that ingredients other than soy globulins dont
contribute to the network structure 72.
12/04/2014 12:04:47.
3.3 Emulsification
globulin, ovalbumin and -casein was found to increase with heat denaturation, whilst that
of -lactolobulin and bovine serum albumin decreased. Kato et al 75b found the curvilinear
correlation between the foaming power of proteins and surface hydrophobicity during heat
denaturation, and found no significant correlation between the foam stability and the
surface hydrophobicity of proteins.
Because of the low surface hydrophobicity, large molecular size, and low molecular
flexibility, glycinin cannot adsorb rapidly to the air-water interface 76. Liu et al 77 isolated
acidic subunits from 11S, and found that isolated acidic subunits adsorbed to the air-water
interface faster than 11S.
Rivas et al 78 found that 7S formed a stronger film at interface than 11S irrespective of pH
or addition of NaCl, and interpreted that the 7S globulin molecules had a greater degree of
intra- and inter-molecular cohesion and so they formed more ordered films. Their view was
consistent with generally accepted concept that molecules with more available
hydrophobic residues develop stronger, and more concentrated, gel-like structures at an
interface, since hydrophobic interactions contribute more rigidity to the film.
Wagner et al 76b showed that the dissociation, deamidation and reduction of glycinin led to
a decrease in molecular size, and increase in surface hydrophobicity and electric charge
thus improving the emulsifying function.
Kimura et al 79 suggested that carbohydrate moieties in 7S globulin plays an important role
in increasing the emulsifying property based on the comparative study of 7S and 11S
globulins extracted from pea, faba bean, cowpea and French bean.
It has been known that in soybean seeds there are oil bodies consisting of a triacylglycerol
core, which is covered by a layer of phospholipids and a protein oleosin. One very special
12/04/2014 12:04:47.
characteristics of soy proteins is its high oil holding capacity as is seen when tofu-curd is
cooked in hot water. While fat is exuded out when most animal meat is cooked in water, no
oil is exuded out from tofu-curd. Guo et al 25b stated that lipid incorporation took place by
the conjugation of the lipid and protein particles based on the examination of lipid in three
fractions, floating, particulate, and soluble fractions obtained by centrifugation. The oil
bodies are believed to exist in naturally emulsified state, and many research groups have
already extracted from soybeans in aqueous environment without using organic solvent
such as hexane, and studied the application in food products in place of emulsified soybean
oil, for example, in dressings, sauces, dips, beverages, and desserts. Additional advantages
of using natural soybean oil bodies in foods, rather than emulsified bulk soybean oil, are
that neither emulsifiers nor homogenization procedures are required 80.
A review on protein-stabilised emulsions 81 and a review on emulsion stabilisation by
protein-polysaccharide complexes 82 have recently been published.
Jambrak et al 70 reported that both emulsifying and foaming ability of soy proteins can be
increased by ultrasound treatment. Ultrasound treatment with 20 kHz probe increased the
solubility of soy protein concentrates, the specific surface area and EAI.
Chove et al 86 showed a possibility to modify the structure by microfiltration.Fractionation
was carried out on SPI produced by isoelectric precipitation of a crude protein extract, and
retentates and permeates were obtained.Emulsions stabilised by the retentates exhibited
higher emulsion stability index (ESI) and emulsifying activity index (EAI) than those
stabilised with permeates. They also found that the fractions exhibiting high functionality
in terms of solubility, foaming and emulsifying properties were also richer in 7S globulin
soy protein subunits based on SDS-PAGE.
Chen et al 87 examined the effect of oxidation on the emulsifying properties of soy protein
isolate, and reported that emulsions stabilized by moderately oxidized SPI had a smaller
droplet size and better thermal stability in comparison with the control and over-oxidized
SPI.
Although soybean oil bodies have great advantages as previously noted, they are unstable
to aggregation because of the relatively weak electrostatic repulsions between them over a
wide range of pH values (3<pH<7), and the stabilization of oil body emulsions was studied
by coating with pectin 88 or carrageenans 89. Iwanaga et al 88 showed that citrus pectin-
coated oil bodies had similar or improved stability compared to uncoated oil bodies. Wu et
al 89 using , , - carrageenans to see the effect of electric charge and conformation on the
stabilizing effect, and found that -carrageenan was most effective and attributed it to the
most densely charged helical structure, and thus most effective at creating highly charged
interfacial membranes.
View Online
Heating solutions of the conjugates prior to emulsification was shown to improve their
emulsification properties
Soy soluble polysaccharide (SSPS) has been shown to be a good emulsifier in the food
industry 95, and it was shown recently to be able to prevent the destabilization of SPI
dispersions and SPI-based oil-in-water (O/W) emulsions under acidic conditions 96.
Emulsion gels were reviewed recently by Dickinson97. Since many processed foods can be
regarded as protein-based oil-in-water emulsions which can be converted into soft-solid-
like materials by common food processing operations such as heating, acidification, and
enzyme action, there have been many studies on emulsion gels. Matsumura et al 98 studied
the effect of oil content and droplet size on rheological properties of emulsion gels of 11S
globulin. They found that both storage and loss moduli of the gels increased with
increasing oil content up to 15% oil and with decreasing droplet size. Kim et al 99
confirmed similar results including large deformation behaviour; both breaking stress and
strain increased with increasing oil content (<volume fraction 0.3) and with decreasing
droplet size.
Tang et al 100a reported the preparation and characterization of cold, gel-like soy protein
emulsions consisting of untreated and preheated SPI at a protein concentration of 6% (w/v),
and various oil volume fractions (0.2-0.6) and NaCl concentrations (0-500 mM) by
microfluidization. Both untreated and preheated SPI emulsions were found to behave gel-
like rheologically, but the latter ones were found more gel-like at a comparable oil volume
fractions. Since both steady shear viscosity and storage modulus of the emulsion increased
with increasing oil volume fraction, they attributed it to enhanced inter-droplet interactions,
and they confirmed this bridging flocculation of oil droplets by CLSM observation.
View Online
They dont call this state a gel because the chain-like connected oil droplets may be
essentially a liquid just like a previously so-called weak gel which is essentially a liquid,
and should not be called a gel. There must be a critical oil content above which the
emulsion gel should not be called a gel but should be called gel-like emulsion. Gels will be
broken when subjected to a large deformation while the gel-like emulsion may not be
broken and flow 27c.
Tang et al 101b recently studied transglutaminase-set soy globulin-stabilized emulsion gels,
and found that the increase in the glycinin/-conglycinin ratio progressively increased the
gel stiffness, but significantly decreased the water holding capacity. Their CLSM showed
that increasing glycinin content led to the formation of emulsion gel network with a more
inhomogenous and porous microstructure. They suggested this gel-like emulsion may be
useful as carriers for heat-labile ingredients with health effects.
4. Concluding Remarks
In spite of enormous efforts, the perfect crystal of the main storage proteins glycinin and -
conglycinin has not yet been obtained, which makes difficult to get a clear understanding
of their behaviour. It is urgently required to establish this, to exploit further utilisation of
soybean proteins. It is also necessary to elucidate the interaction among main proteins,
glycinin, -conglycinin, and whey proteins as well as lipids to develop further application
of soy beans because it is not economical to use each extracted/separated protein fractions
in food industry. It will be useful to compare the functional properties, gelling and
emulsification with globular proteins of the other origins to get more insight. It is also
12/04/2014 12:04:47.
required to get systematic understanding about the mechanism of the interaction between
soybean protein and polysaccharides to develop their further the utilisation in food and
pharmaceutical area.
Acknowledgements
Authors wish to thank Prof.T.Ono, Prof.T.Nagano, and Dr.K.Kohyama.
View Online
REFERENCES
(1) Kitamura, K. All About Soybeans; 1st ed.; Science Forum: Tokyo, 2010.
(2) Nishinari, K. In Gums and Stabilisers for the Food Industry 4; Phillips, G. O.,
Wedlock, D., Williams, P. A., Eds.; IRL Press: Oxford, 1988.
(3) Chen, Y. M.; Yamaguchi, S.; Ono, T. J. Agric. Food Chem. 2009, 57, 3831.
(4) Watanabe, T.; Ebine, H.; Ota, T. Soybean Foods; Kohrin: Tokyo, 1991.
(5) Smith, A. K.; Rackis, J. J.; Isnardi, P.; Cartter, J. L.; Krober, O. A. Cereal Chem. 1966,
43, 261.
(6) (a) Iwabuchi, S.; Yamauchi, F. J. Agric. Food Chem. 1987, 35, 205(b) Koshiyama, I.;
Kikuchi, M.; Fukushima, D. J. Agric. Food Chem. 1981, 29, 340(c) Rackis, J. J.;
Wolf, W. J.; Baker, E. C. In Nutritional and toxicological significance of enzyme
inhibitors in foods; Friedman, M., Ed.; Plenum Pub. Corp: New York, 1986.
(7) (a) Pearson, A. M. In Developments in Food Proteins-2 Hudson, B. J. F., Ed.;
Applied Science Pub.: Essex, England, 1983(b) Sorgentini, D. A.; Wagner, J. R. J.
Food Biochem. 1999, 23, 489.
(8) Saio, K.; Kamiya, M.; Watanabe, T. Agricultural and Biological Chemistry 1969, 33,
1301.
(9) (a) Badley, R. A.; Atkinson, D.; Hauser, H.; Oldani, D.; Green, J. P.; Stubbs, J. M.
Biochimica Et Biophysica Acta 1975, 412, 214(b) Peng, I. C.; Quass, D. W.;
Dayton, W. R.; Allen, C. E. Cereal Chem. 1984, 61, 480.
(10) Utsumi, S.; Matsumura, Y.; Mori, T. In Food Proteins and TheirApplications;
Damodaran, S. P., A., Ed.; Dekker: New York, 1997.
(11) (a) Maruyama, N.; Adachi, M.; Takahashi, K.; Yagasaki, K.; Kohno, M.; Takenaka,
12/04/2014 12:04:47.
Y.; Okuda, E.; Nakagawa, S.; Mikami, B.; Utsumi, S. Eur. J. Biochem. 2001, 268,
3595(b) Adachi, M.; Kanamori, J.; Masuda, T.; Yagasaki, K.; Kitamura, K.;
Mikami, B.; Utsumi, S. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 7395.
(12) Ren, C. G.; Tang, L.; Zhang, M.; Guo, S. T. J. Agric. Food Chem. 2009, 57, 1921.
(13) Samoto, M.; Maebuchi, M.; Miyazaki, C.; Kugitani, H.; Kohno, M.; Hirotsuka, M.;
Kito, M. Food Chem. 2007, 102, 317.
(14) Wu, S. W.; Murphy, P. A.; Johnson, L. A.; Fratzke, A. R.; Reuber, M. A. J. Am. Oil
Chem. Soc. 1999, 76, 285.
(15) Deak, N. A.; Murphy, P. A.; Johnson, L. A. J. Food Sci. 2006, 71, C413.
(16) Hu, S. D.; Liu, H.; Qiao, S. Y.; He, P. L.; Ma, X.; Lu, W. Q. J. Agric. Food Chem.
2013, 61, 4406.
(17) Damodaran, S.; Kinsella, J. E. J. Agric. Food Chem. 1982, 30, 812.
(18) Utsumi, S.; Damodaran, S.; Kinsella, J. E. J. Agric. Food Chem. 1984, 32, 1406.
(19) (a) Maruyama, N.; Salleh, M. R. M.; Takahashi, K.; Yagasaki, K.; Goto, H.; Hontani,
N.; Nakagawa, S.; Utsumi, S. J. Am. Oil Chem. Soc. 2002, 79, 139(b) Utsumi, S.;
Maruyama, N.; Satoh, R.; Adachi, M. Enzyme and Microbial Technology 2002, 30,
284.
(20) Hashizume, K.; Watanabe, T. Agricultural and Biological Chemistry 1979, 43, 683.
(21) Guo, J.; Yang, X. Q.; He, X. T.; Wu, N. N.; Wang, J. M.; Gu, W.; Zhang, Y. Y. J.
Agric. Food Chem. 2012, 60, 3782.
(22) Kuipers, B. J. H.; van Koningsveld, G. A.; Alting, A. C.; Driehuis, F.; Voragen, A. G.
J.; Gruppen, H. Food Biophysics 2006, 1, 178.
(23) Tsumura, K.; Saito, T.; Tsuge, K.; Ashida, H.; Kugimiya, W.; Inouye, K. Lwt-Food
Science and Technology 2005, 38, 255.
(24) Jung, S.; Murphy, P. A.; Johnson, L. A. J. Food Sci. 2005, 70, C180.
View Online
(25) (a) Ono, T.; Katho, S.; Mothizuki, K. Biosci. Biotechnol. Biochem. 1993, 57, 24(b)
Guo, S. T.; Ono, T.; Mikami, M. J. Agric. Food Chem. 1999, 47, 901(c) Yuan, Y.;
Wan, Z. L.; Yin, S. W.; Yang, X. Q.; Qi, J. R.; Liu, G. Q.; Zhang, Y. Lwt-Food
Science and Technology 2013, 50, 657.
(26) Ren, C. G.; Tang, L.; Zhang, M.; Guo, S. T. Food Hydrocolloids 2009, 23, 936.
(27) (a) Doi, E. Trends in Food Science & Technology 1993, 4, 1(b) Clark, A. H. In
Functional properties of food macromolecules (2nd Ed.); Hill, S. E., Ledward,
D.A., & Mitchell, J.R., Ed.; Aspem Publishers, 1998(c) Djabourov, M.; Nishinari,
K.; Ross-Murphy, S. B. Physical Gels from Biological and Synthetic Polymers;
Cambridge University Press: Cambridge, 2013.
(28) Dobson, C. M. Nature 2003, 426, 884.
(29) Chiti, F.; Dobson, C. M. Nature Chemical Biology 2009, 5, 15.
(30) (a) Adamcik, J.; Mezzenga, R. Macromolecules 2012, 45, 1137(b) van der Linden, E.;
Sagis, L. M. C. Langmuir 2001, 17, 5821.
(31) Akkermans, C.; Van Der Goot, A. J.; Venema, P.; Gruppen, H.; Vereijken, J. M.; Van
Der Linden, E.; Boom, R. M. J. Agric. Food Chem. 2007, 55, 9877.
(32) Tang, C. H.; Wang, C. S. J. Agric. Food Chem. 2010, 58, 11058.
(33) Nilsson, M. R. Methods 2004, 34, 151.
(34) Bikbov, T. M.; Grinberg, V. Y.; Schmandke, H.; Chaika, T. S.; Vaintraub, I. A.;
Tolstoguzov, V. B. Colloid and Polymer Science 1981, 259, 536.
(35) Nakamura, T.; Utsumi, S.; Mori, T. J. Agric. Food Chem. 1984, 32, 349.
(36) Veerman, C.; Sagis, L. M. C.; van der Linden, E. Macromolecular Bioscience 2003, 3,
243.
12/04/2014 12:04:47.
(37) (a) Nishinari, K. Colloid and Polymer Science 1997, 275, 1093(b) Nishinari, K.
Reports on progress in Polymer Physics in Japan 2000, 43, 163.
(38) Nishinari, K.; Kohyama, K.; Zhang, Y.; Kitamura, K.; Sugimoto, T.; Saio, K.;
Kawamura, Y. Agricultural and Biological Chemistry 1991, 55, 351.
(39) Nagano, T.; Mori, H.; Nishinari, K. J. Agric. Food Chem. 1994, 42, 1415.
(40) Nagano, T.; Akasaka, T.; Nishinari, K. Biopolymers 1994, 34, 1303.
(41) Wang, C. H.; Damodaran, S. J. Agric. Food Chem. 1991, 39, 433.
(42) Mills, E. N. C.; Huang, L.; Noel, T. R.; Gunning, A. P.; Morris, V. J. Biochimica Et
Biophysica Acta-Protein Structure and Molecular Enzymology 2001, 1547, 339.
(43) Nagano, T.; Akasaka, T.; Nishinari, K. Food Hydrocolloids 1995, 9, 83.
(44) Iwabuchi, S.; Watanabe, H.; Yamauchi, F. J. Agric. Food Chem. 1991, 39, 27.
(45) Alexander, M.; Dalgleish, D. G. Colloids and Surfaces B-Biointerfaces 2004, 38, 83.
(46) Nik, A. M.; Alexander, M.; Poysa, V.; Woodrow, L.; Corredig, M. Food Biophysics
2011, 6, 26.
(47) Ringgenberg, E.; Alexander, M.; Corredig, M. Food Hydrocolloids 2013, 30, 463.
(48) Nagano, T.; Mori, H.; Nishinari, K. Biopolymers 1994, 34, 293.
(49) Renkema, J. M. S.; Gruppen, H.; van Vliet, T. J. Agric. Food Chem. 2002, 50, 6064.
(50) Wongprecha, T.; Takaya, T.; Kawase, T.; Nagano, T.; Nishinari, K.; Nishinari, K.,
Ed.; Elsevier Science
Amsterdam, 2000.
(51) Lakemond, C. M. M.; de Jongh, H. H. J.; Paques, M.; van Vliet, T.; Gruppen, H.;
Voragen, A. G. J. Food Hydrocolloids 2003, 17, 365.
(52) Foegeding, E. A.; Kuhn, P. R.; Hardin, C. C. J. Agric. Food Chem. 1992, 40, 2092.
(53) Kohyama, K.; Nishinari, K. Food Hydrocolloids 1992, 6, 263.
(54) Renkema, J. M. S.; van Vliet, T. Food Hydrocolloids 2004, 18, 483.
(55) Kohyama, K.; Sano, Y.; Doi, E. J. Agric. Food Chem. 1995, 43, 1808.
View Online
(56) (a) Yoshida, M.; Kohyama, K.; Nishinari, K. Biosci. Biotechnol. Biochem. 1992, 56,
725(b) Kohyama, K.; Nishinari, K. J. Agric. Food Chem. 1993, 41, 8.
(57) Deman, J. M.; Deman, L.; Gupta, S. Food Microstructure 1986, 5, 83.
(58) Liu, H. H.; Kuo, M. I. J. Texture Stud. 2011, 42, 1.
(59) Nagano, T. Sen-I Gakkaishi 2009, 65, P462.
(60) (a) Nio, N.; Motoki, M.; Takinami, K. Agricultural and Biological Chemistry 1986,
50, 851(b) Motoki, M.; Seguro, K. Trends in Food Science & Technology 1998, 9,
204.
(61) (a) Yasir, S. B. M.; Sutton, K. H.; Newberry, M. P.; Andrews, N. R.; Gerrard, J. A.
Food Chem. 2007, 104, 1491(b) Tang, C. H.; Wu, H.; Chen, Z.; Yang, X. Q. Food
Res. Int. 2006, 39, 87.
(62) Nonaka, M.; Sakamoto, H.; Toiguchi, S.; Yamagiwa, K.; Soeda, T.; Motoki, M. Food
Hydrocolloids 1996, 10, 41.
(63) Dumoulin, M.; Ozawa, S.; Hayashi, R. J. Food Sci. 1998, 63, 92.
(64) Molina, E.; Defaye, A. B.; Ledward, D. A. Food Hydrocolloids 2002, 16, 625.
(65) Puppo, C.; Chapleau, N.; Speroni, F.; de Lamballerie-Anton, M.; Michel, F.; Anon,
C.; Anton, M. J. Agric. Food Chem. 2004, 52, 1564.
(66) Maltais, A.; Remondetto, G. E.; Gonzalez, R.; Subirade, M. J. Food Sci. 2005, 70,
C67.
(67) Maltais, A.; Remondetto, G. E.; Subirade, M. Food Hydrocolloids 2009, 23, 1647.
(68) Speroni, F.; Anon, M. C. Food Hydrocolloids 2013, 33, 85.
(69) Nakao, Y.; Yamaguchi, T.; Taguchi, T. Nippon Shokuhin Kogyo Gakkaishi 1994, 41,
141.
12/04/2014 12:04:47.
(70) Jambrak, A. R.; Lelas, V.; Mason, T. J.; Kresic, G.; Badanjak, M. Journal of Food
Engineering 2009, 93, 386.
(71) Hu, H.; Fan, X.; Zhou, Z.; Xu, X. Y.; Fan, G.; Wang, L. F.; Huang, X. J.; Pan, S. Y.;
Zhu, L. Ultrasonics Sonochemistry 2013, 20, 187.
(72) Yoshimura, M.; Naito, S.; Nagano, T.; Nishinari, K. J. Jpn. Soc. Food Sci. Technol.-
Nippon Shokuhin Kagaku Kogaku Kaishi 2007, 54, 143.
(73) Dickinson, E. In Handbook of Hydrocolloids; Phillips, G. O., Williams, P. A., Eds.;
Woodhead Publ, 2009.
(74) Pearce, K. N.; Kinsella, J. E. J. Agric. Food Chem. 1978, 26, 716.
(75) (a) Kato, A.; Nakai, S. Biochimica Et Biophysica Acta 1980, 624, 13(b) Kato, A.;
Osako, Y.; Matsudomi, N.; Kobayashi, K. Agricultural and Biological Chemistry
1983, 47, 33.
(76)(a) Kinsella, J. E. J. Am. Oil Chem. Soc. 1979, 56, 242(b) Wagner, J. R.; Gueguen, J. J.
Agric. Food Chem. 1995, 43, 1993(c) Wagner, J. R.; Gueguen, J. J. Agric. Food
Chem. 1999, 47, 2181.
(77) Liu, M. X.; Lee, D. S.; Damodaran, S. J. Agric. Food Chem. 1999, 47, 4970.
(78) Rivas, H. J.; Sherman, P. Colloids and Surfaces 1984, 11, 155.
(79) Kimura, A.; Tandang-Silvas, M. R. G.; Fukuda, T.; Cabanos, C.; Takegawa, Y.;
Amano, M.; Nishimura, S. I.; Matsumura, Y.; Utsumi, S.; Maruyama, N. J. Agric.
Food Chem. 2010, 58, 2923.
(80) (a) Iwanaga, D.; Gray, D. A.; Fisk, I. D.; Decker, E. A.; Weiss, J.; McClements, D. J.
J. Agric. Food Chem. 2007, 55, 8711(b) Chen, Y. M.; Ono, T. J. Agric. Food Chem.
2010, 58, 6485.
(81) Lam, R. S. H.; Nickerson, M. T. Food Chem. 2013, 141, 975.
(82) Evans, M.; Ratcliffe, I.; Williams, P. A. Current Opinion in Colloid & Interface
Science 2013, 18, 272.
View Online
1 INTRODUCTION
With an expected growth of the world population to above 9 billions (2050) and a
subsequent increasing demand for high nutritional foods, there will be an enormous
pressure on the future food production system to fulfil this demand. The protein supply is
in this respect most critical, not only for human consumption, but also for feed
applications. Plant proteins are known to be more sustainable and cost effective than
12/04/2014 12:04:49.
animal proteins1,2. (Partly) replacing animal protein in existing products with (new) plant
protein ingredients or developing new plant protein-based products is an effective
approach in making more protein available for a larger part of the world population. In
addition to the use of proteins directly from crops, large organic side and waste streams are
available for the extraction of proteins, also in the food-chain.
Currently, the majority of the plant proteins used in food and feed are plant-storage
proteins purified from seeds, where the most abundant protein in the world is located in the
green leaves. This protein, Ribulose-1,5-bisphosphate carboxylase oxygenase, or most
commonly known by the shorter name RuBisCO, is an enzyme that catalyzes the first
major step of carbon fixation, a process by which atmospheric carbon dioxide and water is
converted to energy-rich molecules such as glucose, using sunlight3. In green parts of
plants, the protein RuBisCO can make up to 50% of the total protein fraction.4 Besides its
abundant presence throughout the world, RuBisCO is of potential interest for human
consumption due to its exceptional high nutritional value. This can be illustrated by the
high amount of essential amino acids, which closely match the recommendations by the
FAO/WHO (in 2011) for three target groups based on their age and their nutritional needs
(Table 1). Especially lysine and the sulphur-containing amino acids, which are regarded as
critical in low-meat diets are present at high concentrations in RuBisCO.
View Online
2 Extraction of RuBisCO
But proteins do more. Besides the nutritional aspect, proteins contribute to textural and
sensorial properties of the food product. Changing a current processed food composition is
therefore a very complex process, balancing parameters such as product liking (taste,
12/04/2014 12:04:49.
3. Techno-functional properties
3.1. Gelling properties
RuBisCO protein isolate is able to form a gelled network at lower protein concentrations
compared to soy and whey proteins (Figure 1). These results are in agreement with results
in literature regarding the gelling behavior 10,11 in which RuBisCO was extracted at lab-
scale. The results show that, at pH 70, the heat induced gelation of RuBisCO starts at
around 2% w/w protein in water, while for whey proteins it only starts around 10% w/w
proteins and for soy this is as high as 12% w/w proteins. The high gelling potential of
RuBisCO fits application in deserts very well where in many cases it could reduce the
amount of dairy proteins. The high gelling potential of RuBisCO was also shown using the
RuBisCO protein isolated at semi-industrial scale: a self-supporting gel was formed with
4.5% w/w protein at pH 7.0 either in the presence or absence of salt. These gels did not
show any syneresis over the course of hours at room temperature. RuBisCO forms a
coarse/particulate microstructure: the structure becomes less dense by the addition of salt
during the gelling but the increase in coarseness does not seem to affect the stability of the
gel and no syneresis is appearing within 24 hours.
12/04/2014 12:04:49.
250
Foam Volume (mL)
200
150
100
50
0
0 5 35
Time (min)
Figure 2: Foam volume over two hours of RuBisCO extracted at lab scale (white) and at
semi-industrial scale (light grey), soy (black) and whey protein (dark grey) isolates at two
different pH (4.5: filled bars, and 7: pattern bars).
12/04/2014 12:04:49.
material. However, the foam volume and foam stability at pH 7.0 was still better than those
of whey and soy proteins.
Figure 3: Left: Visual observation and Right: CLSM micrograph of RuBisCO fibre
produced using NIZO protocol.
12/04/2014 12:04:49.
3. CONCLUSIONS
RuBisCO shows very promising functional properties as a food ingredient. Its high gelling
and foaming makes it very applicable in many food products. Combining this with the
healthy aspects, its high protein efficiency and the enormous abundance around the world
in plants and algae, it has the potency to become a major source of protein in future diets.
The next step is a fine tuning of the industrial process to improve the purity and the colour
of the final isolate; moreover we are investigating the potential of the sides streams arising
from our process, such as the the fibre (polysaccharide) stream, as well as the green stream
(rich in polyphenols and chlorophyll).
References
1. Aiking, H., J. de Boer and J. Vereijken, Sustainable protein production and
consumption: Pigs or peas?, ed. Springer, Dordrecht, 2006, The Netherlands.
2. Steinfield, H., P. Gerber, T. Wassenaar, V. Castel, M. Rosales and C. de Haan, FAO
report, Rome, ISBN 92-5-105571-7, 2006, Lifestocks long shadow, Environmental issues
and options.
3. Weissbach, A.; Horecker, B. L.; Hurwitz, J. J. Biol. Chem., The enzymatic formation
of phosphoglyceric acid from ribulose diphosphate and carbon dioxide,1956, 218, 795.
4. Ellis, J.R. Trends Biochem. Sci., The most abundant protein in the world, 1979, 4,
241.
5. Barbeau, W. E. and Kinsella, J.E. Food Rev. Int., Ribulose biphosphate
carboxylase/oxygenase (RuBisCO) from green leaves Potential as a food protein, 1988,
4, 93.
6. Bickoff, E.M., D. de Fremery, R.H. Edwards, B.E. Knuckles, G.O. Kholer and R.E.
Miller, US patent application , Preparation of soluble edible protein from leafy green crops,
1976, 3,959,246.
View Online
7. Bickoff, E.M., D. de Fremery, R.H. Edwards, B.E. Knuckles, G.O. Kholer and R.E.
US patent application, Miller Preparation of soluble edible protein from leafy green crops,
1977, 4,006,078.
8. Johal, S. US patent application, Preparation and crystallization of fraction I protein
from plant sources, 1982, 4,334,024 (published 1982-06-08).
9. Wildman, S.G. and P. Kwanyuen, US patent application, Process for isolation of
ribulose 1,5-diphosphate carboxylase from plant leaves, 1981, 4,268,632.
10. Knuckles, B. E. and Kohler, G.O. J. Agric. Food Chem., Functional properties of
edible protein concentrates from alfalfa, 1982, 30, 748-752.
11. Sheen, S. J. and Sheen, V. L. J. Agric. Food Chem., Functional Properties of Fraction
I from Tobacco Leaf, 1985, 33, 79-83.
12/04/2014 12:04:49.
PROTEIN-POLYSACCHARIDE INTERACTIONS: PHASE BEHAVIOUR AND
APPLICATIONS
1 INTRODUCTION
Due to the low entropy gain during mixing of the large molecules, protein and
polysaccharide are rarely miscible. Thermodynamically compatible mixtures only exist
under specific conditions such as at very low concentrations, or when the two components
are chemically and structurally very similar. The interactions between protein and
polysaccharide determine their phase behaviour.9 Attractive interaction generally leads to a
phase separation called associative phase separation where the protein and polysaccharides
are condensed in the same phase leaving the other phase depleted (Figure 1a). The driving
forces for associative phase separation are the electrostatic interactions between
polysaccharide and protein and an entropic gain by releasing counterions upon formation
of protein-polysaccharide complexes. It has been shown that pH, ionic strength, protein-
polysaccharide ratio (Pr/Ps), total biopolymer concentration, charge density and
conformations of protein and polysaccharide are important parameters controlling
associative phase separation.1 Associative phase separation is observed between oppositely
View Online
S S
Pr Ps Pr Ps
(a) (b)
Figure 1
Typical phase diagrams of associative type (a) and segregative type (b). S,
Pr and Ps stand for solvent, protein and polysaccharide, respectively. The dotted lines
represent the tie lines of phase compositions.
charged proteins and polysaccharides, and more generally between negatively charged
polysaccharides carrying groups such as sulphate or carboxyl and positively charged
proteins such as at pHs below their isoelectric points (pIs). Repulsive interaction generally
leads to a phase separation called segregative phase separation where the protein and
polysaccharide are enriched into two separate phases (Figure 1b). The driving force for
segregative phase separation are polymer-solvent and polymer-polymer interactions as
depicted by Flory-Huggins theory and therefore the major factors influencing them are
12/04/2014 12:04:51.
temperature and polymer concentration. Segregative phase separation often occurs between
uncharged polysaccharide and protein, or similarly charged polysaccharide and protein.
Turgeon and the co-workers elegantly summarized the factors that influence protein-
polysaccharide interactions and phase separations.5 They pointed out that the combination
of the intrinsic and external parameters, as mentioned above, dictate the mechanism of
phase separations, that is, nucleation and growth vs. spinodal decomposition. Nucleation
and growth mechanism normally leads to a dispersed micro-phase structure during phase
separation, whereas spinodal decomposition leads to a bi-continuous micro-phase structure.
These microstructures can be frozen, deformed and aligned by applying thermally induced
gelation and shearing, creating rich options for food structural design.5, 10
of the protein that would repulse polysaccharides or a too low pH leads to the protonation
of the acidic groups that would eliminate the electrostatic attraction. In a few cases,
complex coacervation was also observed at pHs above the pI of the protein. This has been
explained by the presence of positive charge patches on the protein surface despite the
overall net charge being negative.14
12/04/2014 12:04:51.
Figure 2Construction of phase diagram for bovine serum albumin-sugar beet pectin
complexes: (top) determination of characteristic pHs, pHc and pH, from light scattering
intensity I90 (circle), hydrodynamic radius Rh (square) and turbidity Wat550 nm (triangle)
for Pr/Ps = 1; (bottom) phase diagram established in pH-Pr/Ps coordinate based on pHc
(circle) and pH (square), and pHd (triangle). The phase diagram encompasses five
regions, (I)-(V) and was highlighted with different patterns indicating different emulsion
stabilities when stabilising oil-in-water emulsions: dotted pattern, stable emulsions;
stripped pattern, unstable emulsions showing reversible flocculation; shadowed pattern,
unstable emulsions showing irreversible flocculation induced by droplet bridging. The
horizontal dashed line represents the isoelectric point of BSA and the vertical dotted line
represents the maximum stoichiometry of BSA/SBP.
View Online
below pHc. This however is now questionable, particularly in the pH region below the
isoelectric point of the protein and above pHc, that is, Region II as shown in Figure 2b.
This is the pH region where the protein and polysaccharide are already oppositely
charged and therefore they should have been electrostatically associated with each
other. By probing in this particular region using GPC-MALLS, our recent study did
find that even in Region II protein-polysaccharide complexes are already present,
which was assigned to intramolecular soluble complexes.18 The structural transitions of
protein-polysaccharide complexes were re-proposed in relation to the phase diagram
(Figure 3): Region I, individual soluble polymers; Region II, stable intramolecular
soluble complex; Region III, intermolecular soluble complex; Region IV,
intermolecular insoluble complex and Region V, a second region for individual soluble
polymers.
(2) With decreasing Pr/Ps ratio r, pHc first slightly falls and then drops dramatically. The
inflexion point (marked by the vertical dotted line) indicates a maximum stoichiometry
of BSA bound to SBP, and is supported by zeta potential measurements.18, 20 The
vertical dotted line therefore divides the phase diagram into the polysaccharide-
excessive region (to the left) and the protein-excessive region (to the right). The
definition of polysaccharide or protein-excessive regions has important implication in
stabilising oil-in-water emulsions using protein-polysaccharide complexes.21
(3) In the BSA-excessive region, pHc is slightly above the isoelectric point of BSA. This
means that the complexation can happen even when the protein is slightly negative.
This can be attributed to the presence of positive patches on the surface of BSA due to
uneven distribution of charges.9, 14 BSA can interact with SBP through the positive
patches even if its overall net charge is negative. It should be pointed out that in the
case of more hydrophobic proteins such as sodium caseinate, hydrophobic aggregation
between proteins can promote the formation of protein-polysaccharide complexes,
leading to pHc being more conspicuously higher than the pI of the protein.19
View Online
Associative phase separation has been widely applied to the food, cosmetic,
pharmaceutical, medical and biotechnological industries to purify proteins/macromolecules,
encapsulate active ingredients, and to structure and stabilize food products etc.1 Protein-
polysaccharide complexes represent a powerful tool to modulate the rheological and thus
textural properties of food systems, ranging from viscous to viscoelastic and even a gelling
behaviour. The behaviour depends on the conformation of the polysaccharides used and
the extent of the attractive interaction between protein and polysaccharide. For example,
12/04/2014 12:04:51.
of the components undergoes gelation. This is the so-called viscoelastic phase separation
and it can result in various microstructures.35 In the study of the gelatin/dextrin system, the
development of microstructures was found to strongly depend on the quench temperature
and thus the gelation of the gelatin phase.36 The coupling of segregative phase separation
with gelation produced quite different mechanical properties, which is sensitive to thermal
history.37
When the polysaccharide participating in segregative phase separation has a high
polydispersity, the situation can be much more complex than the phase separation of
monodisperse systems. Molecules varying in size and shape react differently toward
segregation with the other component, which leads to molecular fractionation of the
polydisperse component accompanying phase separation.38 Phase separation induced
molecular fractionation has been utilized to fractionate and concentrate functional
component from polydisperse materials, such as gum arabic.39 Figure 5 shows a typical
phase diagram of polydisperse mixtures containing gum arabic (GA) and sugar beet pectin
(SBP). GA is a highly polydisperse and heterogeneous polysaccharide, containing also
some proteinaceous material. It consists of three major fractions, namely, arabinogalactan-
protein complex (AGP), arabinogalactan (AG) and glycoprotein (GP). The AGP is a high
molecular weight fraction and is responsible for the emulsifying functionality of GA. The
cloud points determined by visual observation significantly deviate from the binodal points
determined from phase composition analysis. This seemingly contradictory phenomenon
has been explained as an indication of phase separation-induced molecular fractionation.38
The reason is that the cloud points were determined largely based on single phase mixtures,
whereas the binodal points were based on phase separating mixtures. Molecular
fractionation during phase separation changes the constitution of the original polydisperse
material and therefore leads to the difference in critical phase separation concentrations
measured by the two methods.
12/04/2014 12:04:51.
(B)
Figure 5 Phase diagram of gum arabic-sugar beet pectin system: cloud points ()
determined by visual observation and binodal points determined by phase composition
analysis for mixed systems with two different SBP starting concentrations: (, ) 0.8%
SBP and (, ) 1.0% SBP. The solid and open symbols represent the upper and lower
phase, respectively. The broken lines are tie-lines. Reproduction from reference 39.
View Online
The molecular fractionation in GA-rich phase was illustrated in Figure 6a and 6b.The
peak 1 in GPC-MALLS elution profiles (Figure 6a) corresponds to the AGP fraction, and
has been normalized. The peak 2 corresponds to the AG and GP fractions in gum arabic.
At a fixed SBP concentration of 1.0%, the peak 2 decreases with increasing GA
concentration, indicating a relative decrease in the AG and GP fractions and an increase in
the AGP fraction. This clearly demonstrates a molecular fractionation of GA during phase
separation with SBP. Interestingly, it was found that the fractionation efficiency (defined
as AGP content) is linear to the extent of phase separation (defined as the phase volume
ratio of upper phase to lower phase) in a double logarithmic plot, regardless of initial phase
separating concentration of SBP. Practically, this method can be used to improve the AGP
View Online
content of gum arabic up to three times the original content. The resultant materials have
an enhanced emulsifying functionality in stabilising oil-in-water emulsions, compared with
the control gum arabic.39
Associative and segregative phase separation could convert to each other and under certain
experimental conditions can even coexist, which leads to much more complex phase
behaviours and microstructures. Figure 7 is the microstructure of -carrageenan/fish
gelatingels formed by mixing the two biopolymers at high temperature followed by cooling
below the conformational ordering temperature of -carrageenan.40 The microstructure
constitutes of droplets dispersed in a background that underwent further phase separation,
which is a result of segregative phase separation of the two biopolymers on top of a
preceding associative phase separation. The droplets should arise from the complex
coacervation of -carrageenan/fish gelatin and the phase separated background should be
attributed to the segregative phase separation induced by the conformational ordering of -
carrageenan. The coexistence of the two types of interactions and phase separation is
possibly in a kinetically trapped state, due to the ordering and gelation of -carrageenan.
The conversion and coexistence of different types of interactions and phase separation
have been studied in detail using pigskin gelatin and -carrageenan.9 The phase diagram
was established in an ionic strength-temperature coordinate (Figure 8). Depending on the
combination of the experimental parameters, four different phase behaviours could occur,
including compatible, segregative phase separation, associative phase separation, and
12/04/2014 12:04:51.
coexistence of both types of the phase separations. The different phase behaviours created
distinctly different microstructures, which are expected to lead to significant difference in
mechanical properties.9 Moreover, the phase boundaries were found to closely correlate
with the conformational orderings of -carrageenan and gelatin. It is therefore the coupling
of conformational orderings that brings such complex phase behaviours.
3 CONCLUSION
The large spectrum of available protein-polysaccharide combinations together with diverse
types of molecular interactions and phase behaviour provide numerous possibilities to
obtain the desired effects when designing food microstructures, textures and functionalities.
Advantage could also be taken of the molecular characteristics of proteins and
polysaccharide, such as conformational transitions, gelation and aggregation, and the
processing conditions, such as thermal treatment and shearing, to further widen the design
toolbox, leading to innovative applications of protein-polysaccharide mixtures.
Acknowledgements
The authors acknowledge the financial support from the National Natural Science
Foundation of China (31171751), the Program for New Century Excellent Talents in
University (NCET-12-0710), the Key Project of Chinese Ministry of Education (212117),
the Project from the Ministry of Human Resources and Social Security of China, the Key
Project of Natural Science Foundation of Hubei Province (2012FFA004) and the
University Innovative Team Project from Hubei Provincial Department of Education
(T201307).
References
1 C. Schmitt, C. Sanchez, S. Desobry-Banon and J. Hardy, Crit. Rev. Food Sci. Nutr.,
1998, 38, 689.
2 C. Schmitt and S. L. Turgeon, Adv. Colloid Interface Sci., 2011, 167, 63.
View Online
1 INTRODUCTION
(W) is primarily directed towards fracture (Wf) for particulate gels. For stranded gels
the applied energy is either elastically stored (Ws) or dissipated (Wd). The energy
balance can then be denoted as W= Ws+Wd+Wf.1 The fore mentioned mechanical
properties have been shown to relate to texture perception that may go from
spreadable to crumbly for particulate and stranded networks, respectively.2,3
Understanding of how to control the energy balance with regard to microstructure and
texture perception is of key importance for the food industry to modulate their
products towards desired sensory properties, especially when new (alternative)
protein sources are involved.
Texture is to a large extent determined by the properties of the structural elements
and their mutual interactions. These structural aspects stem from the aggregation
behaviour of the individual proteins, which in turn is determined by the molecular
characteristics and their ability to interact during processing. Subsequently, an
assembled microstructure may consist of molecules (nm) and protein molecules that
are assembled into flexible fine stranded structure elements (0.1-5 m), and coarse
stranded or particle shaped structure elements (5-50 m). At any length scale, protein
(structure elements) can be subjected to (food grade) chemical or enzymatic
modification to tune their function in a spatial network. Up to now, this was however
only done for proteins at a molecular level. 4-6
Functionalization of specific structure elements and their interactions is a tool in
understanding which length scales are relevant for tuning texture and breakdown
properties. The type, shape and dimensions of these structure elements determine the
efficiency and gel strength of the established spatial network. Hence, this work
sketches the potential of different types of structure elements made from gelatin, whey
protein and soy protein to direct macroscopic behaviour. On a molecular scale,
View Online
modulation of gelatine is performed to alter its assembly into fine stranded networks
and the subsequent macroscopic breakdown behaviour. Modification of whey protein
is performed on an aggregate level to show the efficiency of thiolation of different
supramolecular structures (fibrillar and amorphous aggregates) with regard to gelation
propensity. On a microstructural level, particulate soy protein networks are tuned
through the presence of calcium salts for their fracture behaviour. We show that
control over texture and macroscopic properties can be obtained by modulation of
protein functionality at different levels of protein organization.
For fine-stranded protein gels the structural building blocks are typically linear and
have limited thickness. Gel strength is dominated by the flexibility (thickness) of the
strands, the mesh size of the stranded network, and the interaction energy in the
assembly points of the individual strand making up the spatial structure. The
presence of sugar moieties cross- linked to gelatine molecules may impair with strand
assembly, both kinetically and interaction energy. Recently we have shown that one
can introduce significant numbers of such steric moieties along a gelatine molecule
using the Maillard reaction4 without losing the major functional property of adopting
triple helix configurations, believed to be essential for efficient network formation.
12/04/2014 12:04:53.
Also differences in charge appear to have a limited impact on the propensity and
kinetics of gel formation5. From small deformation rheology (not shown), or when
applying large deformation rheology (Figure 1A) it becomes evident that at a
mechanical level the introduction of steric groups is very effective, as illustrated by the
gradual lower Youngs modulus with increasing degree of modification (DM).
Combined with calorimetric and structural characterization we have been able to
postulate the hypothesis that it is not the number of junctions that is affected by steric
modulation, but the interaction energy of the junctions7.
Figure 1B shows that despite the decreasing Youngs modulus with increasing
DM, the recoverable energy is not affected. Previously it has been shown that
recoverable energy of semi-solid protein gels was directly related to the sensory
attribute crumbliness.8 Evidently from this work it can be concluded that
Maillardation of gelatines are likely not to lead to changes on this sensory property.
The true fracture stress and strain (Figure 1C and D), do again follow the same
dependence to DM as the Youngs modulus. These rheological properties have been
related to the number of particles formed during for example oral processing and
thereby affecting indirectly the texture attribute spreadability, an important pre-
requisite for perceptions like creamy.8 As for the melting behaviour of the
Maillardated gelatines the enthalpic changes (results not shown) follow the reduced
fracture stress, while the transition temperatures are not affected, it can be
concluded that already limited degrees of modification will affect this latter texture
attribute, while evaluation up to the fracture point might be preserved.
The example given above illustrates how food-grade modification of gelatines
provides important consequences for the energy balance in gelatine gels. Maillardation
has apparently no significant effect on the stored energy (Ws), as the recoverable
energy is not affected by DM. The fracture energy (Wf) is gradually reduced with DM.
The impact of Maillardation on the dissipated energy (Wd) is not clear at the moment
View Online
and is difficult to assess. It could be speculated that ligated sugar moieties facing
entrapped serum in the network would increase the viscous flow within the pores and
thereby would increase the friction contribution. To substantiate this a more in-depth study
would be required.
It has been documented that the type and shape of aggregates formed in a gelling process
determine the efficiency and gel strength of the established network. Typically, the critical
(minimum) gelling concentration is lower for fibrillar aggregates compared to more
amorphous ones. Amorphous aggregates (micrometer scale) and fibrillar aggregates
(nanometer scale) can be assembled into a spatial network using cold-set gelation, e.g.
using GDL (glucono- -lactone). Whereas for amorphous aggregates this has been widely
studied9,10, for fibrils this is not the case. Due to their differences in dimensions and hence
the interactions between the structure elements, gelation propensity and network properties
will vary significantly between amorphous and fibrillar aggregates.
Intermolecular crosslinking has appeared to be very relevant in determining gel
strength. Typically, when additional functional groups were introduced that would enhance
intermolecular chemical crosslinking higher gel strengths were observed. Alternatively,
when chemical reactivity was suppressed weaker gels were obtained.10 An effective way to
increase the potential of chemical cross-linking between structure elements, is the
introduction of chemical reactive thiol groups. Thiolation of whey protein on a molecular
level was shown to affect the aggregation mechanism or propensity and might affect
thereby the development and final functionality of the formed network.5,6,11 Recently we
View Online
have circumvented this latter issue by modifying whey protein on an aggregate level
(either fibrillar or amorphous aggregates) introducing a (limited) number of thiol groups
per structural building block using S-acetylmercaptosuccinic anhydride (S-AMSA) as
previously described12. Thiol groups were successfully linked to both types of aggregates,
as the number of thiol groups increased drastically (see Table 1). As a result, the chemical
reactivity was boosted significantly by the thiolation, as established from the SEI-index.13
These two types of aggregates with and without modification were used as building blocks
to form a spatial network. From Figure 2A it can be seen that the concentration dependence
of G is almost doubled by introducing additional chemical reactivity. Also for amorphous
aggregates (see Figure 2B) G increases as a result of thiolation although the effect for
fibrillar aggregates is far more pronounced. Moreover, at similar protein concentration,
fibrillar aggregates form much stronger gels than amorphous aggregates. From this work it
can be suggested that, in addition to tuning the spatial dimensionality of the building block,
the network forming properties can be further boosted by thiolation, especially when the
building blocks have a more anisotropic structure. In this way, protein interaction on a
microstructural level can be tuned very efficiently. In doing so, gel properties like firmness
or fracture stress can be easily tuned bearing a direct impact on the afore discussed energy
balance.
Table 1: Number of thiol groups (mM/mM protein) for native, unmodified (U)
and modified (SX) fibrillar (F) and amorphous (A) aggregates of whey protein, as
determined by OPA assay
12/04/2014 12:04:53.
Figure 2: The storage modulus (G) as a function of the protein concentration for A)
fibrillar (F) aggregates and B) amorphous (A) aggregates either unmodified (U) or
thiolated (SX)
View Online
2.3 Selecting the right ions to balance fracture properties and water holding
capacity
Serum release or water expelled from the protein network during compression (e.g.
during oral processing) contributes to dissipated energy in the energy balance
equation. Opposite to serum release is the extent to retain water or water holding,
which in literature is generally accepted to be set by the network (micro)structure.
Hence, following the energy balance, the microstructure determines both mechanical
properties and release of water.14
However, we postulate that the microstructure as such is not the determining factor for
the water holding capacity of soy protein gels. To prove this, particulate soy protein
networks are tuned on a microstructural level through the presence of different
calcium salts and compared for their microstructure, mechanical properties and
water holding capacity. Water holding capacity is measured using an adapted method
of Kocher and Foegeding15 in which the sample is placed in a micro-centrifuge and
the speed of centrifugation is increasing stepwise (20-1000g for 10 min.). This
approach allows us to not only to determine the amount of water remaining in the
gel after centrifugation but also the kinetics of water expulsion as a function of
centrifugational force applied.
Heat-set soy protein gels were prepared from native soy protein isolate (SPI) in
the presence of two calcium salts (CaCl2 and CaSO4) as a trigger for protein
aggregation. Whereas CaCl2 is highly soluble in water, CaSO4 only has a limited
solubility at room temperature. Heating in the presence of the two calcium salts
12/04/2014 12:04:53.
resulted in particulate soy protein networks (as confirmed by CLSM) with different
microstructure (see Figure 3) in terms of pore size and strand thickness/density.
These differences can be partially explained by the fact that the soluble calcium from
CaCl2 binds to the protein instantly while the poorly soluble calcium from CaSO4 is
released more slowly during heating creating a much more dense network.
Differences in the mechanical properties of the gels can be directly related to the
microstructure. The more dense network formed with CaSO4 gives a firmer gel
with a higher Youngs modulus and higher fracture stress than gels triggered by with
CaCl2 (see Table 2). With regard to the energy balance, for CaSO4 the energy is
more directed towards fracture instead of storage as is indicated by the lower
recoverable energy and lower fracture strain (not shown). However, despite the
differences in microstructure and mechanical properties, the water holding capacity (at
maximum centrifugation speed) for the two protein networks is very comparable (see
Table 2). A relevant factor appeared to be the kinetics at which water is released
from the gel as CaCl2 induced gels loose water faster upon applied g-force than
CaSO4 induced gels. This corresponds directly with the more dense structure of the
CaSO4 gels.
View Online
CaCl2 CaSO4
Youngs modulus (kPa) 10.3 0.02 16.6 3.5
Fracture stress (kPa) 4.5 0.3 13.7 0.3
12/04/2014 12:04:53.
3 CONCLUSION
References
1 T. van Vliet and P. Walstra, in Food Colloids and Polymer, Stability and
Mechanical Properties. (Royal Society of Chemistry, Cambridge, 1993), 175-190
2 L. v.d. Berg, T. van Vliet, E. van der Linden, M.A.J.S van Boekel and F. v.d.
Velde Food Biophysics 2008 3, 198.
3 L. v.d. Berg, A.L. Carolas, T. van Vliet, E. van der Linden, M.A.J.S van Boekel
and F. v.d. Velde Food Hydrocoll 2008 22, 1404.
4 K. Broersen, A.G.J. Voragen, R.J. Hamer and H.H.J. de Jongh, Biotechnology
and Bioengineering 2004, 86, 78.
5 E.D. Strange, V.H. Holsinger and D.H. Kleyn, J. Agric.Food Chem. 1996,
44, 54 6 S.C. Kim, N.F. Olson and T. Richardson, Milchwissenschaft 1990, 45,
627.
7 D. Baigts Allende and H.H.J de Jongh, submitted to Food Biophysics (2013)
8 H.H.J de Jongh, Food Biophysics 2011, 6, 303.
9 A.C. Alting, R.J. Hamer, C.G. de Kruif, M. Paques and R.W. Visschers, J Agric
Food Chem, 2003, 51, 3150.
10 A.C. Alting, M. Weijers, M.A. Cohen Stuart. R.J. Hamer, K.C.G. de Kruijf, and
R.W. Visschers, J. Agric.Food Chem. 2004, 52, 623.
11 K. Broersen, A.M.M. Van Teeffelen, A. Vries, A.G.J. Voragen, R.J. Hamer,
12/04/2014 12:04:53.
1 INTRODUCTION
Perennial ryegrass, Lolium perenne L., contains a significant amount of fructan and is
becoming increasingly attractive as a biorefinery feedstock for the production of ethanol
and platform chemicals.1,2 Unlike inulin which consists of linear (2,1) chains of fructose
12/04/2014 12:04:55.
units terminating with a glucose molecule, fructans from perennial ryegrass consist of
fructose chains containing both (2,1) and (2,6) linkages (Figure 1) and vary in chain
length from 3 to ~90 units.3
There has been increasing interest in the use of inulin in food products in recent years
because of its ability to form gels at high concentrations and also because it functions as a
dietary fibre.4 The major commercial source is chicory and it has been estimated that over
350k tons are sold annually.5
Ryegrass may offer an alternative source of fructan and this work aims to characterise
and show structural information important to understanding the similarities and differences
with chicory inulin.
2.1 Materials
Ryegrass fructan samples were obtained by an extraction process developed at IBERS.
Inutec N10 and Hydrolysed Inutec H25P chicory inulin were obtained from BENEO-Bio
Based Chemicals, Belgium. Potassium bromide, sodium chloride and acetonitrile were
obtained from Fisher Scientific and the 2,5-dihydroxybenzoic acid from Sigma Aldrich.
2.2 Methods
For characterisation by FTIR spectroscopy, approximately 1mg of the sample was milled
with approximately 100mg of dried KBr using an agate mortar and pestle for several
minutes until fully mixed to form a very fine powder. The powder was then compressed
into a thin transparent KBr pellet using a 15 Ton Manual Press and a P/N 03000 13mm
pellet die (Max load 10.0 Tons) from Specac Limited. FTIR spectra were recorded using a
Perkin-Elmer FTIR spectrometer spectrum RX 1. 16 scans were performed between 400
and 4000 cm-1. Spectral analysis and display were performed using the interactive Read-
IR3 version3.0 software.
The molecular mass distribution was obtained by Gel Permeation Chromatography (GPC)
using a GE Healthcare Superose 12 GL column in 0.1M NaCl eluent, with Wyatt DAWN
DSP light scattering and Wyatt Optilab DSP refractive index detectors in series. Samples
were introduced through a 200l injection loop after passing through a 0.45m syringe
filter. The molecular weight was determined using Astra for Windows 4.90.08 QELSS
12/04/2014 12:04:55.
2.XX. The Debye model was used for all evaluation analyses. A value of 0.131 was used
for the refractive index increment (dn/dc).7
The IR spectra for the inulin and ryegrass fructan samples are presented in Figure 2 and it
is noted that they are very similar. The spectra display a broad peak from 4000-3000 cm-1
due to O-H stretching. The peak at 2900 cm-1 is assigned to C-H stretching and the peak at
1620 cm-1 is caused by a C=O stretch. Differences are more pronounced in the fingerprint
region (below 1000 cm-1) which could arise from structural variations. At 930 cm-1 and
865 cm-1 ryegrass fructan displays split peaks, while the inulin has single peaks at these
wavelengths. The inulin also has a peak at 780 cm-1, not seen in the ryegrass fructan
sample.
View Online
As noted above, inulin is predominately a (2,1) linear chain whereas ryegrass fructan has
(2,1) and (2,6) linkages and branching. The branching which occurs in the ryegrass is
likely to be the cause of the slight differences observed.
The spectra for the samples obtained by MALDI-TOF mass spectrometry are presented in
Figures 3, 4 and 5. The pattern of individual chain length increases of regular 162 mass
12/04/2014 12:04:55.
units (i.e. a fructose subunit) can be observed in all of the spectra. In Figure 3 one can see
the fructose chain length of the H25P inulin increasing from a degree of polymerisation,
DP, of 2-17, whereas for the N10 inulin it is DP 3-18 and for ryegrass fructan the sugar
chain is increasing, starting at DP 3 and, following to the end, reaches DP 31. This
indicates that the ryegrass sample has a much higher total DP than the inulin samples.
From the data shown the majority of the inulin is mainly DP 3-8 and the ryegrass DP 3-7,
corresponding to a molecular mass of ~500-1400 Da. The noticeable difference in the
spectra for the ryegrass sample is the split peaks (highlighted), not present in the inulin
which may be as a result of chain branching.
Figure 5: MALDITOF mass spectrum of ryegrass fructan with split mass peaks highlighted
The RI elution profiles obtained by GPC for the H25P and N10 inulin samples and the
ryegrass fructan sample are shown in Figures 6 - 8 respectively. The molecular mass
values were found to be: H25P, Mn: 1.42 x 103 Mw: 1.84 x 103, N10, Mn: 2.16 x 103 Mw:
3.12 x 103 and ryegrass fructan Mn: 9.09 x 103 Mw: 1.93 x 104. These values are higher
than might be expected from the MALDI-TOF data and may be due to the fact that GPC is
less sensitive at low molecular mass due to a weak light scattering signal. The contribution
of chains with DP 8-18 observed in the MALDI-TOF spectra also need to be taken into
account when calculating an average value. It should be pointed out also that MALDI-TOF
performs measurements on samples in the solid state, whereas for GPC samples are
hydrated and present in aqueous solution.
View Online
Figure 6: GPC refractive index profile of H25P inulin, Mn: 1.42x103 Mw: 1.84x103
12/04/2014 12:04:55.
Figure 7: GPC refractive index profile of N10 inulin, Mn: 2.16x103 Mw: 3.12x103
View Online
Figure 8: GPC refractive index profile of ryegrass fructan Mn: 9.09x103 Mw: 1.93x104
4 CONCLUSION
12/04/2014 12:04:55.
The investigation has found that the FTIR spectrum of ryegrass fructan shows slight
differences to spectra for chicory inulin which may be due to chain branching. Ryegrass
fructan has a higher degree of polymerisation and molecular mass. Perennial ryegrass
offers a viable and sustainable source of fructans as an alternative to chicory for
application in the food industry.
Acknowledgements
References
1. INTRODUCTION
Plantago, belongs to the Plantaginaceae family, is widely distributed around the world.
12/04/2014 12:04:57.
The main compounds in the Plantago are iridoids, flavonoids and polysaccharides.1 Seeds
of Plantago are used commercially for the production of mucilage, which is confirmed to
be polysaccharide. Psyllium is the common used name for the plant genus Plantago.
Herein, the terms, psyllium and polysaccharides from seeds of the Plantago family are
used interchangeably in the text.
Many studies have been undertaken on the polysaccharide from seeds of P. afra L., P.
ovata Forsk, P. major L. or P. asiatica L., concerned with their structural characteristics
and biological activities.1-4 The structure of psyllium is complex. It is rich in xylose (Xyl)
and arabinose (Ara). Methods including methylation, periodate oxidation and Smith
degradation, combined with gas chromatography, liquid chromatography, infrared
spectrum, gas chromatography-mass spectrum, are applied to characterize polysaccharide
primary structure.5 Studies show that psyllium is highly branched, and its backbone is
composed of -1,4-linked Xylp. Residues distributed in the branched areas, including
-T-linked Xylp, -1,3-linked Xylp, -T-linked Araf and -1,3-linked Araf, are linked to
the backbone by O-2 or O-3 of -1,4-linked Xylp residues.6-11 Psyllium may contain
glucuronic acid (GlcA)7,8 or galacturonic acid (GlaA)9. The ratio of these residues is
different and dependent on the origin of the materials and preparation methods.
Psyllium has many bioactivities, including immune-regulation12,13, lowering blood
sugar14,15, plasma lipids16 and serum cholesterol17-19. It can be used as a drug carrier for
targetable devices for therapeutic agents or controlled release devices.4,20,21 Its function on
improving intestinal track health is widely accepted. Psyllium increases fecal moisture and
the pH value of defecation, and improves short-chain fatty acid production in the
colon.22-25 These bioactivities, especially the intestinal track health improving function, are
View Online
considered to be related to the high viscous and gelling properties of psyllium. Recent
studies only focus on the viscosity and gelling properties of psyllium.26-29 The bioactivity
of psyllium has recently been reviewed4 and the aim of this article is to summarize the
physicochemical properties, structural characteristics and biological activity.
the molecular weight of psyllium extracted by 0.2% NaCO3 from P. asiatica L. was
1.5106 Da.32 The psyllium from P. asiatica L. investigated by our group had a molecular
weight of 1.8106 Da7, which was similar to other reports32. When the mobile phase was
changed into 0.1M NaNO3, there was little difference in molecular weight.
The main monosaccharides in psyllium are Ara and Xyl. Other monosaccharide,
namely, Rha, Glc, Gal, GlcA and GalA, have also been reported to be present in psyllium.
The ratio between Ara and Xyl is dependent on the extraction method used (Table 1). Van
Craeyveld et al reported that the ratio between Ara and Xyl in psyllium from P. ovata was
20.7: 50.3.33 Guo et al reported that total sugar content of Ara and Xyl in the husk was
84.98%, and the ratio between Ara and Xyl was 21.96:56.72.10 The ratio in extracted
psyllium was different to that in husk. However, this study did not provide details of the
sources of psyllium. Tomado et al reported that acidic psyllium from P. asiatica L which
was extracted using water was composed of Ara, Xyl, GlcA and GalA which were in the
ratio of 4.0:10.8:3.3:0.7, respectively.32 Some O-acetyl groups, about 4.8%, were
determined in the psyllium.32 Psyllium, from P.ovata seeds, was also found to contain
small amount of O-acetyl groups, about 4%.34
12/04/2014 12:04:57.
Table 1: Molecular weight and monosaccharide compositions of psyllium from Plantago family
Molecular
Extraction Monosaccharide compositions (mass percentage or
Source weight
method molar ratio)
Da
Rha:Xyl:Ara:Man:Glc:Gal=4.5:5.5:27.3:- :- :1.0
Cold water -
(molar ratio)
Rha:Xyl:Ara:Man:Glc:Gal=1.0:3.2:7.8:trace:2.4:trace
Hot water -
Aisa et al34 P. ovata (molar ratio)
Oxalic acid and
Rha:Xyl:Ara:Man:Glc:Gal=2.6:6.6:13.6:- :1.0:1.0
ammonium -
(molar ratio)
oxalate
Van Husk (without
Ara (20.7%), Xyl (50.3%), Man (1.1%), Gal (4.8%),
Craeyveld P. ovata solvents -
Glc (2.0%), Rha (1.1%)
et al33 extraction)
Rha (1.5%), Ara (21.96%), Gal (3.76%), Glc
Husk -
(0.64%), Xyl (56.72%), Man (0.40%)
Isolation, Characterisation and Properties of Polysaccharides
Guo et al10 - 0.5M NaOH Rha (0.76%), Ara (24.52%), Gal (1.82%), Xyl
-
extraction (gel) (71.16%)Man (1.74%)
0.5M NaOH
Rha (1.89%), Ara (21.78%), Gal (3.55%), Glc
extraction -
(0.49%), Xyl (68.63%), Man (3.66%)
(soluble)
81
View Online
12/04/2014 12:04:57.
82
1.2.M NaOH Rha (1.93%), Ara (20.70%), Gal (5.36%), Glc
-
extraction (16.30%), Xyl (51.28%), Man (4.43%)
purified fractions. Then, main chain of the native fraction was obtained after
partial acid hydrolysis, and methylaton, 1D and 2D NMR analysis was carried
out again. Results showed that the polysaccharide was a highly branched
arabinoxylan having a -1,4-linked Xylp backbone. It did not contain
1,2,3,4-linked Xylp residues. 1,3-linked Xylp residues were distributed among
branched chains, but not on the backbone.
2.3 Rheological properties of psyllium
Psyllium is high in viscosity and has weak gelling properties. Haque et al
reported dispersions of milled seed husk from P. ovata showed weak gelling
properties which were similar to polysaccharides with rigid and ordered
structures.37 Faraknaky et al reported the dynamic rheology of psyllium which
was extracted from P. ovata by cold water.26 Psyllium solutions were
considered to be weak gels since the storage modulus, G, was larger than the
loss modulus, G, throughout the tested frequency range. Concentrations and
temperature could affect the dynamic rheology of psyllium gels, while all gels
at different pH values showed a typical weak gel properties. Guo et al studied
the gelling properties of an alkali insoluble fraction of psyllium polysaccharide
(referred to as AEG) and in particular investigated the effect of Ca2+ on its
dynamic rheology. It was found that AEG could form a weak gel with a fibrous
12/04/2014 12:04:57.
3 BIOACTIVITIES OF PSYLLIUM
reduced when psyllium was taken. Caecal pH and SCFA were not affected.40
However, our research has shown that fermentation of psyllium from P.
asiatica L. in the colon could increase the production of SCFA and decrease the
pH value at a dose of 0.4 g/kg body weight. Concentrations of acetic, propionic,
and n-butyric acids in mouse colonic content of psyllium treated mice were
significantly higher.24 Some studies showed that psyllium from P. ovata
induced apoptosis in colorectal cancer cells independently of molecular
phenotype.44
3.2 Hyperglycemic, Hypolipidemic regulation and Anti-atherosclerosis
Many studies have shown that psyllium had a distinct effect on reducing
blood sugar and lowing blood lipid activities. Psyllium from P. ovata decreased
glucose absorption significantly in type 2 diabetic patients. At the same time, it
also could reduce total cholesterol, LDL cholesterol, and uric acid
significantly.45 It was reported that psyllium could reduce hyperglycaemia in
type I diabetes via inhibition of intestinal glucose absorption.14 Intestinal
disaccharidease activity and glucose transport in 3T3 adipocytes were not
affected by psyllium absorption. Ziai et al reported psyllium decreased serum
glucose and glycosylated hemoglobin significantly in diabetic outpatients.15
HDL-C increased significantly, while the LDL/HDL ratio was significantly
decreased following psyllium treatment.
Psyllium supplementation, which led to reduction in the android fat to
gynoid fat ratio, and LDL cholesterol, could improve fat distribution and the
lipid profile (parameters of the metabolic syndrome) in an at risk population of
View Online
4. CONCLUSION
ACKNOWLEDGEMENTS
The financial support for this study by National High-tech R&D Program of
China (863 Program)( 2013AA102102), the Key Program of National Natural
Science Foundation of China (No: 31130041), National Key Technology R & D
Program of China (2012BAD33B06), National Natural Science Foundation of
China (No: 31301434), the Program for New Century Excellent Talents in
University (NCET-12-0749), the Project of Science and Technology of Jiangxi
Provincial Education Department (KJLD13004), and Research Project of State
Key Laboratory of Food Science and Technology (SKLF-ZZA-201301,
SKLF-KF-201202), is gratefully acknowledged.
REFERENCES
Recalde, J. Salazar and M. Caslake, Am. J. Clin. Nutr, 2007, 85, 1157-1163.
17. M. de Bock, J. G. Derraik, C. M. Brennan, J. B. Biggs, G. C. Smith, D.
Cameron-Smith, C. R. Wall and W. S. Cutfield, PloS one, 2012, 7, 1-6.
18. B. H. Olson, S. M. Anderson, M. P. Becker, J. W. Anderson, D. B. Hunninghake, D.
J. Jenkins, J. C. LaRosa, J. M. Rippe, D. C. Roberts and D. B. Stoy, J. Nutr., 1997,
127, 1973-1980.
19. R. Sola, E. Bruckert, R. M. Valls, S. Narejos, X. Luque, M. Castro-Cabezas, G.
Domenech, F. Torres, M. Heras, X. Farres, J. Vaquer, J. Martinez, M. Almaraz and
A. Anguera, Atherosclerosis, 2010, 211, 630-637.
20. K. Kumar and B. Kaith, Fibers Polym., 2010, 11, 147-152.
21. B. Singh, V. Sharma, A. Kumar and S. Kumar, Int. J. Biol. Macromol., 2009, 45,
338-347.
22. J. A. Marlett and M. H. Fischer, J. Nutr., 2002, 132, 2638-2643.
23. J. L. Hu, S. Nie, C. Li, Z. H. Fu and M. Y. Xie, J. Agric. Food Chem., 2013, 61,
6092-6101.
24. J. L. Hu, S. P. Nie, F. F. Min and M. Y. Xie, J. Agric. Food Chem., 2012, 60,
11525-11532.
25. J. L. Hu, S. P. Nie and M. Y. Xie, Food Chem., 2012, 138, 2338-2345.
26. A. Farahnaky, H. Askari, M. Majzoobi and G. Mesbahi, J. Food Eng., 2010, 100,
294-301.
12/04/2014 12:04:57.
27. Q. Guo, S. W. Cui, Q. Wang, H. D. Goff and A. Smith, Food Hydrocolloids, 2009,
23, 1542-1547.
28. W. Liu, Z. Xie, B. Zhang, Q. Wang, W. Yao, X. Gao and L. Yu, J. Agric. Food
Chem., 2010, 58, 1615-1621.
29. W. Liu, B. Zhang, Q. Wang, Z. Xie, W. Yao, X. Gao and L. Yu, J. Agric. Food
Chem., 2010, 58, 172-179.
30. S. Al-Assaf, G. O. Phillips, P. A. Williams, S. Takigami, P. Dettmar and M. Havler,
Proc. Nutr. Soc., 2003, 62, 211-216.
31. S. Saghir, M. S. Iqbal, M. A. Hussain, A. Koschella and T. Heinze, Carbohydr.
Polym., 2008, 74, 309-317.
32. Tomoda Masashi, Yokoi Maemi and I. Kazuyo, Chem. Pharm. Bull., 1981, 29,
2877-2884.
33. V. Van Craeyveld, J. A. Delcour and C. M. Courtin, Food Chem., 2009, 112,
812-819.
34. H. A. Aisa, M. Abdurahim, A. Yili, L. Xui, R. K. Rakhmanberdyeva and B. T.
Sagdullaev, Chem. Nat. Compd., 2006, 42, 347-348.
35. J. F. Kennedy, J. S. Sandhu and D. A. T. Southgate, Carbohydr. Res., 1979, 75,
265-274.
36. M. H. Fischer, N. Yu, G. R. Gray, J. Ralph, L. Anderson and J. A. Marlett,
Carbohydr. Res., 2004, 339, 2009-2017.
37. A. Haque, R. K. Richardson, E. R. Morris and I. Dea, Carbohydr. Polym., 1993, 22,
223-232.
38. Z. Cheng, J. Blackford, Q. Wang and L. L. Yu, J. Funct. Foods, 2009, 1, 44-49.
View Online
283-286.
50. J. H. Kim, T. W. Kang and Y. K. Ahn, Arch. Pharmacal Res., 1996, 19, 137-142.
FLAXSEED KERNEL DIETARY FIBRE: PARTIAL STRUCTURE AND
PHYSICOCHEMICAL CHARACTERISATION
Huihuang Ding1,2, Steve W. Cui1,2,*, Qi Wang2, Jie Chen3, Nam Fong Han4, H. Douglas
Goff1
1
Department of Food Science, University of Guelph, Guelph, N1G 2W1, Canada
2
Guelph Food Research Centre, Agriculture and Agri-Food Canada, Guelph, N1G 5C9,
Canada
3
State Key Laboratory of Food Science and Technology, Jiangnan University, Wuxi,
214122, China
4
Natunola Health Inc., Ottawa, K2E 7T8, Canada
1 INTRODUCTION
Flaxseed (Linum usitatissimum L.), one of the most economically important oilseed
12/04/2014 12:04:58.
crops, is rich in soluble and insoluble dietary fibres. Compared to the other cereals and
oilseeds like wheat, barley, oat, soybean, etc.,1 flaxseed has a higher content of dietary
fibres (27~28%), lignan (0.6~1.3%) and minerals (3~4%),2,3,4,5 but a very low content of
starch (<1.6%).6
US National Cancer Institute (NCI) targeted flaxseed as one of the six plant materials
for studying as cancer preventive foods and ingredients.7 Nutritional research on flaxseed
has shown that most of the functional effects of flaxseed can be attributed to its alpha-
linolenic acid (ALA), lignan, and dietary fibre components, among which dietary fibres
play a significant role in protecting against several chronic diseases.
The health benefits of dietary fibre include (but not limited to) protecting against
several chronic diseases (e.g. diabetes, obesity, colon cancer), reducing blood cholesterol
levels, and improving insulin sensitivity, etc.8,9 Due to the high nutritional potential and
relatively low glycemic index (GI) value, flaxseed and its products have been used for
preparing valueadded food products, which are summarized in Table 1.
The other potential products of flaxseed dietary fibres are listed below: (1) Flaxseed
hull: lignan-enriched bakery product, lignan tablet, laxative tablet with high soluble and
insoluble dietary fibre; (2) Flaxseed meal or ground flaxseed: high dietary fibre donut,
bagel, muffin; whole wheat bread and other bakery product with nutty flavour and low
gluten; (3) Whole flaxseed kernel: value-added flaxseed products with high ALA and
dietary fibre (cake, snack bar, salad topping, bakery and roasted product); (4) Flaxseed
gum: stabilizer, water-holding and suspending agent, cake, cheese spread, ice cream, sour
cream, whipping cream, frozen dessert, sauce, etc.; (5) Soluble flaxseed kernel dietary
fibre (flaxseed kernel gum, FKG): low caloric food or beverage, salad dressing, oil and
flavour emulsion, bakery fillings, icing, confectioneries, processed meat product, pie
filling, etc.; (6) Insoluble flaxseed dietary fibre: bulking and laxative agent, drug therapy
for constipation.
View Online
Previous research on flaxseed dietary fibre mainly focussed on soluble flaxseed gum
from whole seed and flaxseed hull.2,22,23,24,25,26,27,28 A recent study in our research group
discovered that the kernel of flaxseed contained about 20% of dietary fibre, which has not
been reported before. De-oiled flaxseed meal, which is a by-product of flaxseed oil
industry and is normally discarded or used as livestock feeds, is abundant in flaxseed
kernel dietary fibres (FKDF). To promote further utilization of flaxseed, investigation on
structure and physicochemical characteristics of FKDF is an important step to explore their
View Online
potential, thus producing more value-added flaxseed products in food, nutrition, and drug
industries.
2 METHODS
2.1 Materials
Flaxseed kernels (~70% purity) were obtained by a patented de-hulling technique29
from Natunola Health Inc. (Ontario, Canada). Pure flaxseed kernels (>99.9% purity) were
collected through further sieving and cleaning. The cultivar of flaxseed was CDC Sorrel.
Figure 1 Extraction and fractionation procedure of flaxseed kernel dietary fibres (FKDF)
The separate anatomic parts of whole flaxseed were observed by SEM in Figure 2a.
Chemical composition of flaxseed kernel (cultivar: CDC Sorrel) is shown in Table 2. There
was a high content of flaxseed oil (about 44%) followed by protein (about 26%), and total
dietary fibre (about 20%). After oil extraction and enzymatic hydrolysis of protein, the
FKDF is mostly in the supporting structure of the cell walls (Figure 2c, 2d). There was
around 72% total dietary fibre in the cell wall, while there was still around 23% protein
left.
12/04/2014 12:04:58.
Figure 2: Scanning Electron Microscopy (SEM) images of whole flaxseed (a), bar
represents 100 m; flaxseed kernel (FK) before oil extraction (b), bar represents 10 m;
and FK after oil extraction (c, d), bar represents 176 m and 20 m, respectively.
Most of FKDF, which are linked through non-covalent and covalent bonds with other
tissue constituents, cannot be directly extracted by hot water. After being released by
chelating agent and alkaline solution from the network of the cell wall, the extracted FKDF
fractions are all water soluble.
(14%~17%).
The linkage patterns of FK-KPI are presented in Table 4. Six main partially
methylated alditol acetate (PMAA) peaks were observed in the GC profile, which were
assigned as 2,3,4-Me3-Xyl (15.43%), 2,3,4,6-Me4-Gal (12.62%),2,3-Me2-Ara (5.23%), 3,4-
Me2-Xyl (9.10%), 2,3,6-Me3-Glc (14.35%), and 2,3-Me2-Glc (25.70%) by MS spectra
indicating the presence of t-Xylp, t-Galp, 5-Araf, 2-Xylp, 4-Glcp, and 4,6-Glcp. Sugar
residues with the molar ratio less than 1% were not included in Table 4. Through further
fractionation, it was found that FK-KPI was composed of two main polymers even though
its PDI values were low (PDI<1.6). The linkage patterns of two fractions of FK-KPI are
currently under investigation, and detailed structures will be reported separately.
View Online
4. CONCLUSION
Extraction results and SEM images revealed that FKDF were mostly in the supporting
structure of the cell walls. Five FKDF fractions were obtained with the molecular weight
ranging from 486 kDa to 1660 kDa. Monosaccharide compositions were different among
FKDF fractions. Alkaline solution extracted FKDF fractions have relatively higher MW
and higher percentage of arabinose. All FKDF fractions showed the ability to reduce the
surface tension of pure water, among which FK-KPI exhibited the best surface activity.
12/04/2014 12:04:58.
The linkage patterns of FK-KPI have been obtained, and detailed structural
characterization is currently under investigation. These findings could be useful for
understanding their functions and may lead to the development of more value-added
flaxseed products.
Acknowledgements
The authors would like to thank Ms. Cathy Wang, Dr. Ying Wu in Agricultural and
Agri-food Canada, and Dr. Sandy Smith in University of Guelph for their technical
assistance.
References
1. Dhingra D., Michael M., Rajput H., & Patil R. T. (2012). Dietary fibre in foods: a
review. Journal of Food Science and Technology 49, 255-266.
2. Cui S. W. & Mazza G. (1996). Physicochemical characteristics of flaxseed gum. Food
Research International 29, 397-402.
3. Johnsson P., Kamal-Eldin A., Lundgren L. N., & Aman P. (2000). HPLC method for
analysis of secoisolariciresinol diglucoside in flaxseeds. Journal of Agricultural and
Food Chemistry 48, 5216-5219.
4. Morris, D. H. (2007). Flax: A health and nutrition primer, 4th ed. Flax Council of
Canada, [cited: Mar. 11, 2013]. Available at:
http://www.flaxcouncil.ca/english/index.jsp?p=primer&mp=nutrition
5. Mueller K., Eisner P., Yoshie-Stark Y., Nakada R., & Kirchhoff E. (2010). Functional
properties and chemical composition of fractionated brown and yellow linseed meal
(Linum usitatissimum L.). Journal of Food Engineering 98, 453-460.
View Online
6. Nutrient Data Laboratory, USDAs National Nutrient Database for Standard Reference,
Release 24, 2011. [cited: Jun. 14, 2012]. Available at:
http://www.ars.usda.gov/nutrientdata
7. Oomah B. D. & Mazza G. (1999). Health benefits of phytochemicals from selected
Canadian crops. Trends in Food Science & Technology 10, 193-198.
8. Health Canada. Policy for labelling and advertising of dietary fibre-containing food
products. (2012). [cited: May 7, 2012]. Available at:
http://www.femaflavor.org/sites/default/files/Health%20Canada's%20Fibre%20Policy_
Feb%2027%202012.pdf
9. Elleuch M., Bedigian D., Roiseux O., Besbes S., Blecker C., & Attia H. (2011). Dietary
fibre and fibre-rich by-products of food processing: Characterisation, technological
functionality and commercial applications: A review. Food Chemistry 124, 411-421.
10. Stewart S. & Mazza G. (2000). Effect of flaxseed gum on quality and stability of a
model salad dressing. Journal of Food Quality 23, 373-390.
11. Qin L., Xu S. Y., & Zhang W. B. (2005). Effect of enzymatic hydrolysis on the yield of
cloudy carrot juice and the effects of hydrocolloids on color and cloud stability during
ambient storage. Journal of the Science of Food and Agriculture 85, 505-512.
12. Xu Y., Hall C. III., & Wolf-Hall C. (2008). Fungistatic activity of heat-treated flaxseed
determined by response surface methodology. Journal of Food Science 73, 250-256.
13. Sinha S. & Manthey F. A. (2008). Semolina and hydration level during extrusion affect
quality of fresh pasta containing flaxseed flour. Journal of Food Processing and
Preservation 32, 546-559.
14. Manthey F. A., Sinha S., Wolf-Hall C. E., & Hall C. A.,III (2008). Effect of flaxseed
12/04/2014 12:04:58.
flour and packaging on shelf life of refrigerated pasta. Journal of Food Processing and
Preservation 32, 75-87.
15. Alpers L. & Sawyer-Morse M. K. (1996). Eating quality of banana nut muffins and
oatmeal cookies made with ground flaxseed. Journal of the American Dietetic
Association 96, 794-796.
16. Chetana, Sudha M. L., Begum K., & Ramasarma P. R. (2010). Nutritional
Characteristics of Linseed/flaxseed (Linum Usitatissimum) and its Application in
Muffin Making. Journal of Texture Studies 41, 563-578.
17. Hall C. A., Manthey F. A., Lee R. E., & Niehaus M. (2005). Stability of alpha-linolenic
acid and secoisolariciresinol diglucoside in flaxseed-fortified macaroni. Journal of
Food Science 70, 483-489.
18. Manthey F. A., Lee R. E., & Hall C. A. (2002). Processing and cooking effects on lipid
content and stability of alpha-linolenic acid in spaghetti containing ground flaxseed.
Journal of Agricultural and Food Chemistry 50, 1668-1671.
19. Shearer A. E. H. & Davies C. G. A. (2005). Physicochemical properties of freshly
baked and stored whole-wheat muffins with and without flaxseed meal. Journal of
Food Quality 28, 137-153.
20. Pohjanheimo T. A., Hakala M. A., Tahvonen R. L., Salminen S. J., & Kallio H. P.
(2006). Flaxseed in breadmaking: Effects on sensory quality, aging, and composition of
bakery products. Journal of Food Science 71, 343-348.
21. Hyvarinen H. K., Pihlava J. M., Hiidenhovi J. A., Hietaniemi V., Korhonen H. J. T., &
Ryhanen E. L. (2006). Effect of processing and storage on the stability of flaxseed
lignan added to bakery products. Journal of Agricultural and Food Chemistry 54, 48-
53.
22. Muralikrishna G., Salimath P. V., & Tharanathan R. N. (1987). Structural Features of
an Arabinoxylan and a Rhamnogalacturonan Derived from Linseed Mucilage.
Carbohydrate Research 161, 265-271.
View Online
23. Cui S. W., Mazza G., & Biliaderis C. G. (1994). Chemical structure, molecular size
distributions, and rheological properties of flaxseed gum. Journal of Agricultural and
Food Chemistry 42, 1891-1895.
24. Oomah B. D., Kenaschuk E. O., Cui S. W., & Mazza G. (1995). Variation in the
composition of water-soluble polysaccharides in flaxseed. Journal of Agricultural and
Food Chemistry 43, 1484-1488.
25. Fedeniuk R. W. & Biliaderis C. G. (1994). Composition and physicochemical
properties of linseed (Linum usitatissimum L.) mucilage. Journal of Agricultural and
Food Chemistry. 42, 240-247.
26. Goh K. K., Pinder D., Hall C. E., & Hemar Y. (2006). Rheological and Light
Scattering Properties of Flaxseed Polysaccharide Aqueous Solutions.
Biomacromolecules 7, 3098-3103.
27. Naran R., Chen G. B., & Carpita N. C. (2008). Novel rhamnogalacturonan I and
arabinoxylan polysaccharides of flax seed mucilage. Plant Physiology 148, 132-141.
28. Qian, K. Y., Cui, S. W., Wu, Y., & Goff, H. D. (2012). Flaxseed gum from flaxseed
hulls: Extraction, fractionation, and characterization. Food Hydrocolloids 28, 275-283.
29. Cui S. W., Han N. F. (2006). Process and apparatus for flaxseed component
separation. U Patent 7,022,363, issued Apr. 4, 2006.
30. Dusterhoft E. M., Voragen A. G. J., & Engels F. M. (1991). Nonstarch Polysaccharides
from Sunflower (Helianthus-Annuus) Meal and Palm Kernel (Elaeis-Guineenis) Meal
Preparation of Cell-Wall Material and Extraction of Polysaccharide Fractions. Journal
of the Science of Food and Agriculture 55, 411-422.
31. Selvendran R. R. & Stevens B. J. H. (1985). Developments in the Isolation and
12/04/2014 12:04:58.
Analysis of Cell Walls from Edible Plants. Brett, C.T.and J.R.Hillman (Ed.).Society for
Experimental Biology Seminar Series, 28.Biochemistry of Plant Cell Walls; Meeting,
Glasgow, Scotland. Cambridge University Press: Cambridge, England; New York,
N.Y., Usa.Illus , 39-78.
32. AOAC method. (2000), Official Methods of Analysis 17th Ed., Association of Official
Analytical Chemists, Washington, D.C.
33. Cui S. W., Wood P. J., Blackwell B., & Nikiforuk J. (2000). Physicochemical
properties and structural characterization by two-dimensional NMR spectroscopy of
wheat beta-D-glucan--comparison with other cereal beta-D-glucans. Carbohydrate
Polymers. 41, 249-258.
34. Du Nouy P. L. (1919). A new apparatus for measuring surface tension. Journal of
General Physiology 1, 521-524.
35. Taylor R. L., & Conrad, H. E. (1972). Stoichiometric depolymerization of polyuronides
and glycosaminoglycuronans to monosaccharides following reduction of their
carbodiimide-activated carboxyl groups. Biochemistry 11, 1383-1388.
36. Ciucanu, I., & Kerek, F. (1984). A simple and rapid method for the permethylation of
carbohydrates. Carbohydrate Research 131, 209-217.
THE EFFECTS OF ULTRASOUND ON THE EXTRACTION, PHYSICOCHEMICAL
PROPERTIES AND ANTIOXIDANT ACTIVITIES OF POLYSACCHARIDE FROM
GANODERMA ATRUM
1. INTRODUCTION
Ganoderma (G.) is a famous medicinal fungus found in the regions of China, Japan,
Korea, and Eastern Russia. G. atrum is one of the most important species of
Ganodermataceae which is abundant in Southern China. The polysaccharide from G.
12/04/2014 12:04:59.
atrum (PSG) has been shown to possess excellent bioactivity1, 2. However, the yield of
PSG is low and the large molecular size of PSG affects its further investigation. Ultrasonic
treatment has been widely used to improve the yield recovery, physicochemical properties
and bioactivity of polysaccharides3. Its powerful sonication can not only serve to mill the
raw material and facilitate swelling and hydration to improve the extraction rate, but also
simply split the most susceptible intramolecular/intermolecular interaction (such as
hydrogen bonding) of polysaccharides to modify the solution properties and chemical
structure4, 5. However, these effects on bioactivity are controversial at present. Some
literature studies reported positive results that ultrasonication could improve the bioactivity
of polysaccharides6, while others found that polysaccharide treated by ultrasonic had
poorer bioactivities7. In order to improve the yield recovery and modify the molecular
structure of PSG, we employed the ultrasonic technique in the present work and assessed
its effects on structure, physicochemical properties and antioxidant activity of PSG.
2.1. Materials
The fruiting bodies of G. atrum were cultivated in Ganzhou, Jiangxi Province, China,
and harvested during the period of July, 2009.
Figure 1: Response surface plots (3D) for the effect of independent variables on the
response of polysaccharide yield. (A) extraction temperature and ratio of water to raw
material, the level of extraction time fixed at zero; (B) extraction time and ratio of water to
raw material, the level of extraction temperature fixed at zero; (C) extraction time and
extraction temperature, the level of ratio of water to raw material fixed at zero.
View Online
Microscopic results showed that the ultrasonic treated residues became much smaller
than those of hot water treated and untreated samples (Fig. 2). This phenomenon could
induce ultrasonic cavitation and facilitate the hydration effect to enhance mass transfer
from the material to solvent at a lower temperature and in a shorter time.
The water solubility test was conducted to investigate the effect of ultrasonic
12/04/2014 12:04:59.
Table 1: The yield, water solubility and chemical composition of PSG by different methods
Water Uronic Sugar component (%)
sampl Yield a a Neutral Protein a
solubility acid
e (%) sugar a (%) a
(%) Man Glc Gal
(g/L) (%)
U-PS 1.030. 7.030. 11.050. 37.7
861.02 62.640.68 12.15 50.08
G 04 23 21 7
H-PS 0.730. 3.040. 10.900. 37.1
752.61 71.120.42 13.39 49.48
G 03 16 26 3
a
: Data were shown as mean SD, n = 3.
Figure 3: The IR spectrum of PSG extracted by different methods (The upper was the
spectrum of H-PSG, and the bottom was the spectrum of U-PSG).
12/04/2014 12:04:59.
Figure 4: The microstructure images of PSG extracted by different methods. (A) and (B)
The AFM images of H-PSG and U-PSG on mica, respectively. Image size: 15x15 m. (C)
and (D) The environment scanning electron micrographs of H-PSG and U-PSG on mica,
respectively. The measurement was under low vacuum mode at magnification 6000x.
Figure 5: Antioxidant activities of H-PSG and U-PSG in different antioxidant models: (A)
scavenging ability of DPPH radicals; (B) inhibition effect on iron-induced liver lipid
peroxidation; (C) reducing power. All experiments were conducted in triplicate, and data
were presented as mean values SD (P < 0.05).
View Online
The iron-induced microsomal lipid peroxidation system was conducted to evaluate the
inhibition effect of PSG. The inhibition rate of U-PSG and H-PSG increased as the
increase of concentration (Fig. 5B). With the concentrations of 0.1 to 1.6 mg/mL, the
inhibition rates were from 4.11 to 57.31% and from 0.35 to 39.50% for U-PSG and H-PSG,
respectively. It has been reported that the presence and involvement of iron ions and
oxygen could induce the generation of free radicals in the iron-induced lipid peroxidations
assay10. The much stronger inhibition ability of U-PSG compared to H-PSG suggested that
ultrasonic radiation increased the binding sites of polysaccharide with metal ions, which
could interfere with the free radical reaction chains.
Fig. 5C showed that both U-PSG and H-PSG exhibited a strong ability for reducing
Fe3+ compared to the positive control. The reducing power of U-PSG was 3.13 at the
concentration of 6.4 mg/mL, which was even higher than Vc at the same concentration.
The reducing properties were generally associated with the presence of reductants, which
have been shown to exert antioxidant action by breaking the free radical chain through
donating a hydrogen atom13. Therefore, the better antioxidant effect of ultrasonic treated
samples might be due to the exposure of more reducing groups.
4. CONCLUSIONS
The present study systematically evaluated the effect of ultrasonic treatment on the
extraction, physicochemical properties and antioxidant activities of PSG. An optimal
12/04/2014 12:04:59.
process by BBD showed that ultrasonic-assisted extraction shortened the extraction time
and decreased the temperature with a better recovery of PSG than HWE. Ultrasonic
treatment had no significant effect on the primary structure of the polysaccharide.
However, it increased the water-solubility and altered the microstructure of PSG.
Furthermore, the antioxidant activities were also compared between U-PSG and H-PSG.
Overall, the present study has shown that the ultrasonic technique could be applied in the
extraction of PSG to improve the extraction efficiency and physicochemical properties as
well as antioxidant activities.
Acknowledgements
The financial support for this study by the Key Program of National Natural Science
Foundation of China (No: 31130041), National Natural Science Foundation of China (No:
31071532, 21265011), National Key Technology R & D Program of China
(2012BAD33B06), the Program for New Century Excellent Talents in University
(NCET-12-0749), and Research Project of State Key Laboratory of Food Science and
Technology (SKLF-ZZA-201301), is gratefully acknowledged.
View Online
References
[1] W. J. Li, S. P. Nie, Y. Chen, Y. X. Wang, C. Li, M. Y. Xie, J. Agr. Food Chem., 2011,
59, 3707.
[2] Q. Yu, S. P. Nie, J. Q. Wang, P. F. Yin, W. J. Li, M. Y. Xie, Int. Immunopharmacol.,
2012, 14, 362.
[3] Z. M. Wang, Y. C. Cheung, P. H. Leung, J. Y. Wu, Bioresource Technol., 2010, 101,
5517.
[4] Y. T. Li, J. Li, S. Y. Guo, H. L. Li, Ultrason. Sonochem., 2005, 12, 183.
[5] M. Vinatoru, Ultrason. Sonochem., 2001, 8, 303.
[6] L. Y. Zhao, Y. H. Dong, G. T. Chen, Q. H. Hu, Carbohyd. Polym., 2010, 80, 783.
[7] M. Zhang, L. N. Zhang, P. C. K. Cheung, V. E. C. Ooi, Carbohyd. Polym., 2004, 56,
123.
[8] Y. Chen, M. Y. Xie, S. P. Nie, C. Li, Y. X. Wang, Food Chem., 2008, 107, 231.
[9] R. Mukerjea, G. Slocum, J. F. Robyt, Carbohydr. Res., 2007, 342, 103.
[10] H. X. Chen, M. Zhang, B. J. Xie, Food chem., 2005, 90, 17.
[11] M. Oyaizu, Jpn. J. Nutr. [Eiyogaku Zasshi], 1986, 44, 307.
[12] S. C. Szu, G. Zon, R. Schneerson, J. B. Robbins, Carbohydr. Res., 1986, 152, 7.
[13] H. M. Qi, Q. B. Zhang, T. T. Zhao, R. Chen, H. Zhang, X. Z. Niu, Z. E. Li, Int. J. Biol.
Macromol., 2005, 37, 195.
12/04/2014 12:04:59.
OPTIMIZATION OF ULTRASOUND-ASSISTED EXTRACTION OF KONJAC FLOUR
FROM AMORPHOPHALLUS MUELLERI BLUME
1 INTRODUCTION
Konjac tubers with outer diameters between 19 and 25 cm, weight 3 kg0,2, were
collected from a konjac farmer at Sumberbenda Village, Saradan district, Madiun Regency,
Indonesia. All the chemicals used were analytical grade, while the water was glass
distilled. Commercial konjac flour (KGM, Made in USA) was bought through on line
trading.
2.1 Sample Preparation
Clean and fresh Amorphophallus muelleri Blume was sliced to 0.5-1 cm thickness and
dried for 11 hours in a cabinet dryer (60 0C). The chips were ground with a stamp-mill
to pass through 30 mesh screens and it was air classified by using a cyclone three times
and filtered to pass through 80 mesh screens. The powder was called crude konjac flour
(CKF). The ultrasound treatment of CKF was then carried out as described in the
experimental design below.
RSM was used to obtain the optimal ultrasonically assisted extraction condition of crude
konjac flour (CKF) from Amorphophallus muelleri Blume to obtain PKF with high
glucomannan content, high viscosity, low calcium oxalate and clear white colour of konjac
flour. The extraction experiment was carried out according to a central composite design
with 2 factors and 5 levels. Two independent variables selected for this paper were
extraction time (min, X1) and solvent/flour ratio (Table 1). For each factor an experimental
range was based on the results of preliminary experiments. Four responses namely,
glucomannan, viscosity, calcium oxalate and degree of whiteness, were chosen as
View Online
Table 1: Independent variables and their levels in the response surface design
k k k 1, k
Y E 0 E i X i E ii X i2 E i, j Xi X j H (1)
i 1 i 1 i 1, j 2
Where, Eo ;Ei ;Eii and Ei.j are regression coefficients in the intercept, linear, quadratic and
interaction terms, respectively; Xi ; Xii ; Xi,j are the independent variables; Y are responses
and is a random error component.
Software of Design Expert 7.1.0 Version Trials (State-Ease, Inc., Minnepolis MN, USA)
was used to obtain the coefficients of the quadratics polynomial model. The quality of the
fitted model was expressed by Lac of fit value, R2 value and Adjusted R2 value and its
12/04/2014 12:05:01.
The irradiation was carried out at 300 dB and 60 kHz frequency.The immersion stainless
steel transducer was a horn type with the length of 6 cm and diameter of 1.8 cm. The
ultrasonic extraction of crude konjac flour (CKF) with Brown Sonic Type 2000 ultrsound
instrument was conducted in three stages of extraction processes. Stage 1, CKF was
dispersed at solvent/flour ratios (ml/g) of 5.17:1, 6:1, 8:1, 10:1, 10.83:1 in 40% ethanol and
the ultrasound treatment was conducted at extraction times of 10.86, 15, 25, 35 and 39.14
minutes. CKF slurry was sieved through a filter paper to separate residue and supernatant.
Stage 2, the residue was again extracted in 60% ethanol, at the extraction time and
solvent/flour ratio as described at stage 1. Finally, the extraction process was repeated in
80% ethanol with processing conditions as stated in stage 1 (Figure 1 and Table 3). The
residue was then dried at 400C in an oven for 12 hours. Each sample and control
(commercial KGM) was stored in airtight containers until it was used for further analysis.
View Online
Filtered
Supernatan
Filtered
Supernatan
Figure 1: Scheme for the extraction of CKF from Amorphophallus muelleri by ultrasound-
assisted extraction procedure
Water content was determined by weight difference after drying the samples, as described
by.36 Ash content was determined gravimetrically.36 Fat content was determined using a
Soxhlet apparatus according to37 Protein content was calculated from the nitrogen content
(N% X 6.25) analyzed by the Kjeldahl method. Starch content was determined by
spectrophotometric method as described by.37 Glucomannan content was determined using
3.5 dinitrosalicylic acid reagent (Sigma-Aldrich) and measured spectrophotometrically at
550 nm as described by.38 Viscosity of 1% konjac flour solution was determined using
spindle needle 1 at 60 rpm of Brookfield Viscometer (Brookfield LD IV) at room
View Online
3.1 Choosing the fitted model for purifying crude konjac flour
The results of analysis variance, lack of fit and the adequacy of the model are summarized
in Table 2. The data showed a good fit with equation (1), which was statistically acceptable
at P<0.05 level and adequate with a satisfactory R2 value (R2 for glucomannan, viscosity
and calcium oxalate were 0.8556, 0.9771 and 0.9311, respectively). Adjusted R2 values for
glucomannan, viscosity and calcium oxalate were 0.7525, 0.9607 and 0.8819, respectively,
indicated that the model for an independent variable on those three responses was a
quadratic model. The fitted quadratic model was observed on antioxidant activity of
polysaccharides from Tremella mesenterica when R2 value and adjusted R2 value were
0.9918 and 0.9812, respectively.41 The fitted quadratic model was also reported by 42,
when R2 value and adjusted R2 value of optimization of the ultrasound-assisted synthesis
of allyl 1-naphyl ether using RSM, were 0.949 and 0.818, respectively.
Furthermore, results of the probability F value for responses (glucomannan, viscosity and
calcium oxalate) were significant (p<0.05). There are only 0.19%, 0.01% and 0.97%
chances, respectively, that the Model F-value for glucomannan, viscosity and calcium
12/04/2014 12:05:01.
oxalate could occur due to noise. On the other hand, probability F value model quadratic
for the degree of whiteness was non significant (p>0.05), and there is 67.35% chances that
noise could occur for the Model F-value. The result of the sequential model sum of the
square of the degree of whiteness for the probability of F-value suggested the model linear
fit for the effect of independent variables on the degree of whiteness with p value<0.05
(0.0255) or 2.55% (data not shown at Table 2).43 also found the model linear fit on
responses of flavanol, total phenol and anthocyanins activity from the fruit pulp of Euterpe
edulis using RSM.
The lack of fit value (Table 2), was over 5% with the model quadratic for responses
such as: glucomannan, viscosity, calcium oxalate, and the model linear for the response of
the degree of whiteness. The model was appropriate, when the lack of fit test has p
value>0.05.44
The principle of ultrasonic-assisted extraction is that ultrasonic waves hit on the vegetal
material cells, break the cells and release any impurities on the surface of the cells as well
as the cells contents into the extraction medium.45 The ultrasonic waves increase the
power and speed of the extraction medium into the material cells and improve the efficacy,
extraction time and yield. 46, 47, 48 The effect of extraction conditions assisted by ultrasound
on responses of PKF is presented in Table 3. As shown in Table 3, a remarkable increase
of the glucomannan and viscosity were observed, when CKF was purified with the
extraction time of 25 minutes and solvent/flour ratio at 8 ml/g. Beyond that time, the range
and solvent/sample ratio had little effect on the glucomannan and viscosity of PKF. A
significant decrease of calcium oxalate was observed over the extraction time range (15
12/04/2014 12:05:01.
39.14 min), calcium oxalate reaching the minimum level of 0.02% at 39.14 min and
solvent/flour ratio 8 ml/g. When the extraction time is short, a lower response on the
degree of whiteness was observed, and the longer the extraction time was beneficial to the
purification process for removing all impurities on the surface of glucomannan granules.
Therefore the maximum score of the degree of whiteness was observed at the extraction
time 39.14 min.
View Online
Responses
Time Solvent/flour
No.a (min) ratio (ml/g) Glucomannan Viscosity Calcium Degree of
(%) (cPs) oxalate whiteness
(%)
1 25.00 5.17 72.18 10300 0.09 60.37
2 35.00 10.00 70.27 9000 0.03 59.30
3 15.00 6.00 75.31 11000 0.09 58.52
4 10.86 8.00 74.18 13000 0.07 58.50
5 25.00 8.00 87.41 14000 0.04 58.97
6 25.00 8.00 86.77 14000 0.05 59.61
7 25.00 8.00 85.97 14000 0.05 59.48
8 25.00 8.00 86.90 14000 0.04 59.72
9 15.00 10.00 74.61 13500 0.08 58.66
10 39.14 8.00 67.70 8500 0.02 63.85
11 25.00 10.83 85.71 13000 0.05 60.04
12 25.00 8.00 79.97 13500 0.05 60.46
13 35.00 6.00 69.81 10000 0.04 60.04
a. Experiments were conducted in a random order.
12/04/2014 12:05:01.
Responses surface methodology using Design Expert 7.1.0. Version Trial resulted in a
predicted second-order polynomial model quadratic for Glucomannan, viscosity and
calcium oxalate, respectively, as follows:
Figure 2: Response surface plots of the glucomannan (a), viscosity (b), calcium oxalate (c)
and degree of whiteness (d) affected by extraction time and solvent/flour ratio.
View Online
To determine optimal levels, the surface plots of the glucomannan content, viscosity,
calcium oxalate and the degree of whiteness affected by extraction time and solvent/flour
ratio were constructed according to Eq. (2), (3), (4) and (5), respectively.
Figure 2a shows the glucomannan content increased slightly as the extraction time
increased, with the optimal extraction time around 25 minutes, and started to decrease as
the extraction time continued. The yield of xyloglucan from apple pomace decreased as the
extraction time increased.49 Meanwhile, the glucomannan content decreased as the
solvent/flour ratio decreased from 10 ml/g to 8 ml/g and reached a minimum at a
solvent/flour ratio of 6 ml/g. Generally, the larger the volume of solvent used to extract a
vegetal material, the higher the yield obtained from the extraction process. 50 This theory
was corroborated that the more extraction solvent used, the higher the total soluble solid
dissolved in the extraction solvent. 51
A plot of viscosity against extraction time and solvent/flour ratio shows a model quadratic
of the optimal extraction condition at around 25 min. and 8 ml/g solvent/flour ratio (Fig
2b). A similar pattern can be seen in the surface plot of calcium oxalate against extraction
time and solvent/flour ratio (Fig. 2c). The yield and purity of polysaccharides from
Zizyphus jujube cv. Jinsixiaozao showing a model quadratic with increasing the yield and
purity in water/solid ratio and temperature ratio. 33
Figure 2d represents the response surface and contour plot of the degree of whiteness.
According to this plot, the fitted model of interaction between the extraction time and
solvent/flour ratio was linear. Fig. 2d suggests that the solvent/flour ratio had no effect on
the degree of whiteness. However, the longer the extraction time the higher the degree of
12/04/2014 12:05:01.
whiteness score, with the optimal condition of the extraction time at around 25 minutes and
8 ml/g solvent/flour ratio. This data was in accordance with Hossain et al. who examined
the effects of amplitude and extraction time on carnosol and apigenin-7-O-glucoside from
marjoram (Origanum majorana) using RSM.52 They reported that the fitted model for
optimizing ultrasound-assisted extraction of antioxidant compounds from marjoram
(Origanum majorana) using RSM was linear.
3.4 Verification
The optimal condition obtained using RSM was as follows: the predicted extraction time,
25.10 minutes and solvent/flour ratio, 8.65:1 ml/g resulted in glucomannan content,
viscosity, calcium oxalate and the degree of whiteness of 85.47%, 13970.7 cPs, 0.044%
and 59.78, respectively. A verification run conducted in duplicate, confirmed the optimal
condition. The real experiment showed that the glucomannan content, viscosity, calcium
oxalate and the degree of whiteness were as follows: 84.37%, 13750.0 cPs, 0.045% and
60.38, respectively. The differences in responses between the predicted results and the real
experiments were less than 5%. The strong correlation between the real and the predicted
results confirm that the response model was adequate to reflect the expected optimization.
The differences between the predicted results and the real, if it was not over 5%, indicated
that the response model was quite accurate.53
In general composition of commercial KGM, in terms of quality, is better than the quality
of the optimal PKF obtained by ultrasound-assisted extraction using RSM (Table 4).
View Online
Table 4: Chemical compositions of the optimal purified konjac flour treatment and
commercial KGM.
Glucomannan of KGM was higher than the optima PKF, but its viscosity was relatively
indistinguishable. The amount of calcium oxalate of the optima PKF was roughly 8 times
higher than the control (commercial KGM), but the degree of whiteness was rather low.
Proximate data of both samples indicated they were somewhat identical, except the starch
content in PKF was twice that in KGM, although, the protein content of PKF was slightly
12/04/2014 12:05:01.
lower than the protein content in KGM. This analytical data showed that the optimization
process of PKF on ultrasonically assisted extraction using RSM, affected positively the
quality of PKF. This remarkable finding highlights that treating olive and olive paste using
ultrasound for 8 min significantly improved the extractability of virgin olive oil.54 These
results are in agreement with a previous study.55
Results for Table 4 conforms the statement of Feng et al., that ultrasonic processing is
establishing itself as a significant food-processing technology with the capability of large
commercial scale-up and good payback on capital investment.56
4. CONCLUSION
The ultrasound-assisted extraction with three stages of ethanol washing for crude konjac
glucomannan flour from Amorphophallus muelleri was performed with two variables (the
extraction time and solvent/flour ratio) based on the RSM. The optimal condition of
extracting crude konjac glucomannan flour was at the extraction time, 25 min. and 6 sec.
and solvent/flour ratio 8.65 ml/g. Under this optimal condition the glucomannan, viscosity,
calcium oxalate and the degree of whiteness were at optimum level. The real experiment
confirmed the optimal condition of the extraction process with p value <0.05. The
experimental results showed that the extraction time and solvent/flour ratio at the centre
and factorial treatments of central composite design were significant to increase the
glucomannan content and viscosity of PKF. The calcium oxalate content was affected by
the extraction time and solvent/flour ratio. The degree of whiteness was significantly
affected by the extraction time, but it was not affected with the solvent/flour ratio.
Proximate analysis and four other parameters (glucomannan, viscosity, calcium oxalate
and degree of whiteness) of the optimal ultrasound treatments of PKF were virtually the
View Online
same as those of commercial KGM. Sonication during the CKF extraction process needs to
be explored on a larger scale.
AKNOWLEDGEMENTS
5. REFERENCES
1 . Widjanarko, S.B., M. Nur., A. Sutrisno and E.S. Murtini. Final report research
project: The establishment of bussiness unit model for non timber. Joint research
between Faculty of Agricultural Technology, University of Brawijaya and Department
of Forestry, East Java Province, Indonesia. Report written in Indonesia. 2006. pp. 1-
100.
2. Zhao, J., Zhang, D., Srzednicki, G., Kanlayanarat, S. and Borompichaichartkul, C. Int.
Food Res. J. 2010. 17:1113-1124.
3. Vidyarathi, I.P. Role of forest in tribal life, In: Sinha, S.P. (Ed.). Tribal and Forest,
Bihar tribal Welfare Research Institute, Ranchi, India. 1987. pp. 23.
4. Sankaran, M., N.P. Singh., C. Datt., M. Nedunchezhiyan., B. Santhosh and S.V.
12/04/2014 12:05:01.
11. Mirastuti, Y.A. The use of ethanol recovery from ethanol waste in the purification
process of konjac flour (Amorphophallus oncophillus). Final report undergraduate
project. Unpublished report, written in Indonesia with abstract in English. 2010. pp. 1-
95.
12. Widjanarko, S.B., A. Faridah and A., Sutrisno. Effect of multi-level ethanol leaching
on physico-chemical properties of konjac flour (Amorphophallus oncophyllus). The
12th Asean Food Conference 2011. 16-18 June, BITEC Bangna, Thailand. OJ- 133.
13. Widjanarko, S.B., A., Faridah and A., Sutrisno. Effect of MultiStage Ethanol
Leaching By Maceration Technique on PhysicoChemical Properties Of Purified
Konjac Flour (Amorphophallus Oncophyllus). The 16th Gums & Stabilizers for the
Food Industry Conference, Wageningen, Netherland: 28th June to July 1st 2011.
14. Sugiyama, N., S. Shimara, and T. Ando. Studies on Mannan and Related Compounds
I. The Purification of Konjac Mannan. 1972. Bulletin Chem. Soc. of Japan. 45:561-56
15. Takigami, S. Konjac Mannan, 25.2.3 Purification of konjac flour, In: G.O. Phillips;
and P.A. Williams, (Eds.). Handbook of Hydrocolloids. 2000. E-Book. CRC Press.
Woodhead Pub. Comp. Cambridge, England.
16. Fang, W. and Wu, P. Food Hydrocolloids. 2003. 18:167-170.
17. Yuan, Z.H., Wu, D.C., Wu, H. and Li. X.H. Studies on ultrasonic wave extracting
method determining konjac glucomannan in konjac refined powder. Zhongguo Zhong
Yao Za Zhi. 2003. 28(4):324-7. Article in Chinese. Abstract in English. Jilin Military
Medical College, Fourth Military Medical University, Jilin 132011, Jilin China.
18. Cheng, L.H., Nur Halawiah, H., Lai, B.N., Yong, H.M., and Ang, S.L. Int. Food Res.
J. 2010. 17:1043-1050.
12/04/2014 12:05:01.
19. Mason, T.J., L. Paniwnyk and J.P. Lorimer. Ultrason. Sonochem. 1996. 3: S253-S260.
20. Wu, J., L. Lidong and F. Chou. Ultrason. Sonochem. 2001. 8: 347-352.
21. Fu, C., T. Haijun, L. Quanhong, C. Tongyi and D. Wenjuan. Ultrason. Sonochem.
2006. 13: 511-516.
22. Ivana, T., M.L. Stanisavljevic, V.B. Lazic and Veljkovic. Ultrason. Sonochem. 2007.
14: 646-652.
23. Ma, Y., J. Chen, D. Liu and X. Ye. Ultrason. Sonochem. 2009. 16 :5762.
24. Penroj. P., Mitchel, J.R. Hill, S.E. and Ganjanagunchorn, W. Carbohydr. Polym. 2005.
59:367-376.
25. Nishinari, K., Williams, P.A. and Phillip, G.O. Food Hydrocolloids. 1992. 6: 199-222.
26. Case, S.E. and Hamann, D.D. Food Hydrocolloids. 1994. 8:147-154.
27. Wu, W.X. Li.K.Y., Zhang, Z.L., Wang, Z.L., Ding, D.N., Jin, J.Q., and Li.Y.L.
Chem.Ind. For. Prod. 1997. 17(2):69-72. In Chinese with English abstract.
28. Nishinari, K and Yoshimura, M. Food Hydrocolloids. 1999. 13: 227-233.
29. Shinsaku, K., Shigetomo, T., Naruhiro, H., and Yoshinori, T. Food Hydrocolloids.
2002. 16(4):289-294.
30. Gao, S., Guo, J., Wu, L and Wang S. Carbohydr. Polym. 2008. 73:498-505.
ScienceDirect.
31. Shukla, P., D. Garai. M. Zafar, K. Gupta and S. Shrivastava. J. Appl. Sci. Environ.
Sanit. 2007. 2(3):93-103.
32. Chang, Y.C., Lee, C.L. and Pan, T.M. Appl. Microbiol. Biotechnol. 2006. 72:654-661.
33. Li, J.W., Ding, S.D. and Ding, X.L. J. Food Eng. 2007. 80:176-183.
34. Silva, E.M. Rogez, H., and Larondelle, Y. Sep. Purif. Technol. 2007. 55:381-387.
35. Zhang, Q.A., Zhang, Z.Q., Yue, X.F., Fan, X.H., Li, T., and Chen, S.F. Food Chem.
2009. 116:513-18.
View Online
8: 299-301.
47. Jacques, R.A., L.S. Freitas, V.F. Perez, C. Dariva, A.P. Oliveira, J.V. Oliveira, and
E.B. Carama. Ultrason. Sonochem. 2007. 14: 612.
48. Paniwnyl, L., H., Cai, S., Albu, T.J. Mason and R. Cole. Ultrason. Sonochem. 2009.
16: 287292.
49. Caili, F., T. Haijun, L. Quanhong, C. Tongyi and D. Wenjuan. Ultrason. Sonochem.
2006. 13: 511-516.
50. Yang, Y. and F. Zhang. Ultrason. Sonochem. 2008. 15: 308-313.
51. Gamse, T. Liquid-Liquid Extraction and Solid-Liquid Extraction. 2002. Institute of
Thermal Process and Enviromental Engineering, Graz University of Technology. pp.
2-24.
52. Hossain, M.B., N.P. Brunton, A. Patras, B. Tiwari, C.P.O. Donnell, A. B.Martin-
Diana, and C. Barry-Ryan. Ultrason. Sonochem. 2012. 19: 582590.
53. Wu, L., R.E. Aluko and H. Corke. J. Cereal Sci. 2006. 44: 117126.
54. Clodoveo, M.L., Durante, V and La Notte, D. Ultrasoni. Sonochem. 2013.
20:1261-1270.
55. Jimmenez, A., G. Beltran, and M. Uceda. Ultrason. Sonochem. 2007. 14:725-731.
56. Feng, H., G.V., Barbosa Canovas and J. Weiss. Ultrasound Technologies for Food and
Bioprocessing. 2001. Springer Publishing, NY, New York.
SOLUTION PROPERTIES OF BRACHYSTEGIA EURYCOMA SEED
POLYSACCHARIDE
1 INTRODUCTION
Water soluble seed polysaccharides have a variety of applications in the Food Industry.
They serve as viscosity modifiers, gelling agents, texturizers, binders and sugar/ice crystal
growth inhibitors. The seed polysaccharides are widely distributed in nature as storage
polysaccharides in the endosperm of many tropical plant seeds mostly of the Leguminosae
12/04/2014 12:05:04.
family. In the seeds, they regulate the physiological functions by water retention by
preventing the seed from completely drying out 1. Polysaccharides from plants such as guar
and locust bean are widely applied industrially because of their viscous characteristics 1, 2.
Brachystegia eurycoma belongs to the family Leguminosae (sub-family:
Caesalpinioideae). The use of Brachystegia eurycoma is currently limited to local culinary
despite promising characteristics of the polysaccharide component of the seed. This is
because there is very little information on the properties of the polysaccharide which
would stimulate expansion in its applications. Available information on Brachystegia
eurycoma seed is mostly centred on the evaluation of the functional application of the seed
flour in foods 3, 4, 5,. The functional applications of Brachystegia seed flour are governed by
the properties of the seed polysaccharide. According to Ikegwu et al. 5, novel foods and
food ingredients must be functionally reliable if they are to be accepted for use. A few
reports on the Brachystegia seed polysaccharide included the investigations on the effect of
its admixture with snail mucin as a wound healing agent 7 and the report on some
rheological and functional properties of the seed gum 8. Anderson et al 9, have studied the
gum exudates from the tree of Brachystegia glaucescens and reported they were acidic and
contained glucuronic acid, 4-O-methylglucuronic acid and galacturonic acid, together with
galactose, minor amounts of arabinose, and relatively high proportions of rhamnose. So far
there is no report on the polysaccharide content, sugar composition, rheological behaviour
in the dilute, semi-dilute and concentrated solutions, and the molecular characteristics of
the polysaccharide of Brachystegia eurycoma seed. In this investigation we report on the
seed composition, gum yield, molecular characteristics, and effect of temperature and
small angle deformation on the properties of the seed polysaccharide. A preliminary report
on some properties of Brachystegia eurycoma seed polysaccharide appeared in our
previous work 10.
View Online
2 METHODS
2.1 Seed composition and extraction of soluble polysaccharide
Whole seeds were selected and their weights determined. These were soaked in water
overnight and the hulls carefully scraped. The endosperms were air dried to constant
weight over a period of one week. The weight of the hull was determined by difference;
the contents of hull and endosperm were reported as percentages of the whole seed.
The endosperm was reduced to coarse bits using a Warring blender, defatted for 12 h with
hexane and spread to dry in a fume chamber. A known weight of sample (~10 g) was
dispersed in 400 ml of water and left in a water bath at 60oC to fully hydrate overnight.
This was blended in a Warring blender, poured into a centrifuge bottle and centrifuged at
2500 rpm for 2 h. The supernatant was collected. The residue was mixed with 200 ml
water and blended and the blender rinsed with water and both transferred into a centrifuge
bottle and centrifuged. The supernatants were pooled, concentrated over a Rotavapor and
polysaccharide precipitated with isopropanol. The polysaccharide was purified by
dispersing in water and reprecipitating with isopropanol. The sample was reconstituted in a
small amount of water and freeze dried.
Technology Coporation, Santa Barbara Ca 93103). The polysaccharide solution (20 ml)
containing 1.297E-3 g/ml (w/w) was subjected to microwave bomb treatment for 40 s to
ensure complete disaggregation 11, filtered through a 0.45 m syringe filter and injected
into a 200 l loop, passed through a combination of Suprema columns (100, 3000 and
30000) packed with 10 m beads of polyhydroxymethacrylate copolymer. The eluent
(0.1M NaNO3 solution containing 0.005% of sodium azide) was pumped (Waters: 515
HPLC Pump, Milford, MA 01757, USA) through a degasser (CSI 6150, Cambridge
Scientific Instruments, England) at a flow rate of 0.5 ml/min. The chromatogram was
analyzed with Astra software with a predetermined dn/dc value of 0.140 ml/g 12.
1 1 1
Fedors Equation: (eq. 1)
[K ]c cm [K ]
1
2(K r 1)
2
View Online
= viscosity, A= pre-exponential factor (Pas), Ea= activation energy for viscous flow
12/04/2014 12:05:04.
The molecular parameters were obtained using the Berry first order polynomial. The
polysaccharide had a number average molecular weight Mn of 3.38 x 105 g/mol and a
weight average molecular weight Mw of 5.4 x 105 g/mol giving a polydispersity index
12/04/2014 12:05:04.
Mw/Mn of 1.60, at mass recovery was 85.5%. The radius of gyration (Rg) was 57.7 nm.
The polydispersity index indicates a fairly wide range of molecular sizes of the
polysaccharide. The molecular parameters are close to Mw of 7.82 x 105 g/mol, Rg of 65
nm and Mw/Mn of 1.7 reported for Copaifera langsdorfii seed polysaccharide 17.
K Kf 1
(eq. 3)
Kq Kf x
1 (W u J ) n
where , o and are the shear, zero shear and infinite shear viscosities, respectively; is
the Cross relaxation time, J is the shear rate and n is the rate index.
The profiles (Figure 3A) show two regions: a zero shear or Newtonian plateau ( = o)
region and a shear thinning or power law region ( ~ J -n). The o increased with increase
in polysaccharide concentration. The zero shear regions covered many decades of shear
rates with shear thinning occurring at fairly high shear rates. The critical shear rate, J crit, at
which shear thinning begins is inversely proportional to the relaxation time ( J crit =1/).
The (1.44E-02 to 3.538E-02 s) was small and increased as polymer concentration
increased indicating very short time scales of molecular restructuring compared with the
time scales of experiment. The viscosity-shear rate profiles are presented as a generalized
flow curve by plotting the reduced viscosity, /o against the Deborah number, J (Figure
12/04/2014 12:05:04.
Figure 4: Zero shear specific viscosity (sp,o) versus volume concentration (c[]) showing :
A) Two concentration regimes; B) Three concentration regimes; C) Data fitted to various
viscosity models.
View Online
The dilute regime had a slope of ~1.0 and crossed the semi-dilute regime at Co = 0.36 g/dL,
where Co corresponds to the onset of coil overlap. The semi-dilute has a slope of 1.9 and
crossed the semi concentrated regime at Cc (0.934 g/dL), the point of completion of
overlap. The intersection between the dilute and semi-concentrated regimes occurred at a
critical concentration, of Ccr of 0.7 g/dL, with a slope of 3.8 in the semi concentrated
regime. The data were also fitted to different models: Martins (eq. 4; km = 0.3) and
expanded Martins for tetrahedral packing (eq. 5; kh =0.4) 22 and modified Kulicke (eq. 6; kh
=0.4, n = 0.5) 23 that can be used to describe the viscosity of polymers over the whole
concentration range (Figure 4C).
All the three models only fitted to the data in the dilute regime but deviated at higher c[].
Similar deviations have been reported for some other polysaccharides at higher
concentrations 22.
A small amplitude oscillation stress sweeps at angular frequency () of 6.253 rad/s was
performed to determine the linear visco-elastic region at the different concentrations.
Figure 5 shows the plot of Gc versus oscillation stress for 3%, 4% and 5% polysaccharide
concentrations.
Gc exhibited a linear relationship in the range of oscillation stress 0.1 to 1.0 Pa at all
concentrations; hence oscillation frequency sweeps were carried out at an oscillation stress
value in this range. Figure 6 shows the variation of Gc and Gs with for 3%, 4% and 5%
polysaccharide concentrations. At low (< 110 rad/s), Gs > Gc and this was observed for
all the concentrations. As increased, both Gc and Gs increased but Gc increased faster
than Gs resulting in a crossover point (Gc = Gs) at of 521.5 rad/s for 3%, 253.5 rad/s for
4% and 110.2 rad/s for 5%; with the crossover point shifting to lower as concentration
increased. In the regions of where Gs > Gc, the solution properties were predominantly
viscous while where Gc > Gs the solution was predominantly elastic.
12/04/2014 12:05:04.
Figure 6: Frequency sweep showing variation of G( opened) and G( closed) with angular
frequency for 3%, 4% and 5% Brachystegia eurycoma polysaccharide solutions
View Online
10
(Pa s); * (Pa s)
0.1
0.01
0.001
0.001 0.01 0.1 1 10 100
(1/s); (rad/s)
1% 1% 2% 2% 3% 3% 4% 4% 5% 5%
12/04/2014 12:05:04.
Figure 8: Effect of temperature on the: A). steady shear viscosity of 4%, B) steady shear
viscosity of 5% aqueous solution Brachystegia eurycoma polysaccharide
View Online
100
5 % Brachystegia 4% Brachystegia
Zero shear viscosity (Pas)
y = 6E-05e3694x
R = 0.9978
10
y = 3E-05e3681.2x
R = 0.998
1
0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036
1/Temperature (K-1)
Figure 9: Arrhenius plots of zero shear viscosity as a function of absolute temperature for
4% and 5% concentrations of Brachystegia eurycoma polysaccharide
A plot of zero shear viscosity against the reciprocal of absolute temperature (Figure 9) was
linear from which the activation energy (Ea) for viscous flow was estimated. At 5%
polysaccharide concentration Ea was 30.7 kJ/mol while at 4% Ea was 14.04 kJ/mol. The
minimum energy required to initiate flow at 5% polysaccharide concentration was more
12/04/2014 12:05:04.
than double that required to initiate the same at 4% concentration. This indicates the
minimum energy to initiate flow was concentration dependent and decreased with
temperature. Similar results of dependence of Ea on concentration and temperature has
been reported for other polysaccharide solutions 19.
100
20 deg C
20 deg C
10 25 deg C
25 deg C
; * (Pa s)
30 deg C
30 deg C
40 deg C
1 40 deg C
50 deg C
50 deg C
60 deg C
60 deg C
0.1
1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03
(1/s); (rad/s)
4 CONCLUSIONS
Brachystegia eurycoma seed endosperm contained 44.3% polysaccharide. The solution
properties indicated the polysaccharide had intrinsic viscosity of 4.3 dL/g and average
molecular weight of 5.4 x 105 g/mol. The polysaccharide solutions exhibited good
viscoelastic properties and showed potential for commercial exploitation.
Acknowledgements
The authors acknowledge the funding support of Leverhulme Trust Foundation to LMN.
References
1. Srivastava, M. & Kapoor, V.P. (2005). Seed galactomannans: an overview.
Chemistry and Biodiversity, Volume 2 (pages 295-317), Verlag Helvertica Chemica
Acta AG, Zurich.
2. Richardson, P.H., Willmer, J. & Forster, T.J. (1998). Dilute solution properties of guar
and locust bean gum in sucrose solutions. Food Hydrocolloids 12, 339-348.
3. Ene-Obong, H.N. & Carnovale, E. (1992). Nigerian soup condiments: Traditional
processing and potential as dietary fibre sources. Food Chemistry 43 (1), 29-34.
View Online
21. Azero, E.G., Lopes, L.L. & Andrade, C.T. (1997). Extraction and solution properties of
the galactomannan from the seeds of Cassia javanica L. Polymer Bulletin 39 (5),
621-625.
22. Masuoka, S. and Cowman, M.K. (2002). Equation of state for polymer solution.
Polymer 43, 3447-3453.
23. Ren, Y., Ellis, P.R., Sutherland, I.W., & Ross-Murphy, S.B. (2003). Dilute and semi-
dilute solution properties of an exopolysaccharide from Escherichia coli strain S61.
Carbohydrate Polymers 52, 189-195.
24. Hwang, J.K. & Shin, H.H. (2000). Rheological properties of chitosan solutions.
Korean-Australian Rheology Journal 12 (3/4), 175-179.
12/04/2014 12:05:04.
STUDIES ON POMELO PECTIN: CHARACTERISATION AND RHEOLOGICAL
PROPERTIES
1 INTRODUCTION
Pectin is an extremely complex polysaccharide typically found in cell walls and middle
lamellae of higher plants. It is composed predominantly of a galacturonic acid backbone
and side-chain of monosaccharides, mainly arabinose and galactose.1 It is widely
employed in food and pharmaceutical applications due to its efficient gelling and
12/04/2014 12:05:06.
stabilising properties. Commercial pectins are typically recovered from citrus peels or
apple pomace as a waste from beverage industries. Other novel sources, including sugar
beets and sunflower heads, have been sought but have not been commercially viable.2-4
Likewise, pomelo (Citrus maxima or Citrus grandis), a citrus fruit native to South East
Asia is another promising source of pectin due to its richness in albedo (spongy white
peel), which accounted for 30% of the fruit weight. So far, only a few studies have been
conducted to extract pectin from pomelo albedo using different extraction techniques. The
pioneering work was conducted by Norziah and co-workers5 in which pomelo pectin was
extracted with sodium hexametaphosphate together with hydrochloric acid at pH 2.2, 75 qC
for 1 h prior to precipitation with different solvents such as ethanol, aluminium salt, and
acetone. Later on, a sequential technique with boiling water (1 h), 1% ammonium oxalate
(pH 6.5, 30 qC, 1 h), and dilute hydrochloric acid (pH 3.5, 100 qC, 1 h) was used.6 More
recently, a much simpler process using hot water (80 qC, 3-5 h) was also performed.7 The
main objective of this study was to explore the alternative extraction of pomelo pectin
using acid treatments. The effects of extraction parameters, such as acid type and pH, on
pectin properties, including its yield, degree of esterification, and average molecular
weight were evaluated. Furthermore, rheological properties of extracted pomelo pectin
were examined.
2 EXPERIMENTAL
1:30 (w/v). The suspension was subsequently adjusted to a specified pH with acid. In
order to attain the final concentration of 1.6 mM, calcium chloride solution was added to
the suspension prior to heating the solution at 80 oC for 120 min. Then, it was cooled
down to a room temperature. After removal of any solid residues by means of filtration,
the filtrate was adjusted to pH 4.5 with 1 M sodium hydroxide and subsequently
concentrated through a rotary evaporator. The concentrated solution was precipitated in
cold ethanol using the volume ratio of ethanol to solution of 4:1. Next, the mixture was
left overnight in order to fully achieve the precipitation of pectin. Afterwards, the
precipitate was separated and purified by means of dialysis (MW cut off 1,000) against
water and was finally lyophilised at -50 oC for 48 h. The entire experiment was carried out
in triplicate prior to further analyses.
Pectin samples were dried in a vacuum oven prior to FTIR-ATR analysis. FTIR spectra
(4000-400 cm-1) were recorded through the Nicolet 6700 FTIR spectrophotometer with a
resolution of 4 cm-1 and 32 scans. Then, the spectra were smoothened and their baselines
were corrected with the built-in software (OMNIC 8) of the spectrophotometer. For the
determination of peak areas at 1735 and 1610 cm-1, which indicate the number of esterified
carboxylic and carboxylic groups, respectively, the peak resolve function and Gaussian
curve fitting analysis within the software were conducted. Pectins with known DE values
of 30, 60 and 90% (Sigma-Aldrich) were used as a standard for making the calibration
curve. The DE of pectin samples was then calculated according to Eq. (1);
12/04/2014 12:05:06.
A1735 (1)
DE (%) u 100
A1735 A1610
where A1735 and A1610 are the peak area at 1735 and 1610 cm-1, respectively.8-9
The molecular weight and its distribution for all studied pectin samples were examined
with gel permeation chromatography (Waters 600E, Waters, USA), coupled with a
refractive index detector. Twenty microliters of pectin solution (0.2% w/v) were injected
into an Ultrahydrogel linear column (7.8 u 300 mm, Waters, USA) with a guard column
(6.0 u 40 mm). An isocratic elution with 0.8 M sodium nitrate was carried out at 30 qC
and a flow rate of 0.6 ml/min. Dextran (MW 4,400-401,000) was used as a standard.
A stock of 1% w/v pectin dispersion was firstly prepared by dispersing pectin in DI water.
This was then stirred using a magnetic bar for 4 h to allow it to completely dissolve.
Subsequently, pectin dispersions at varying concentrations of 0.4-1.0% w/v were prepared
from the stock solution by adjusting it with an appropriate amount of DI water. All
samples were then heated to 50 oC for 10 min before loading onto a rheometer.
View Online
type on pomelo pectin recovery and its properties were explored. Effects of other
extraction parameters including time and temperature on pomelo pectin properties will not
be discussed in this paper but can be referred to from our future publication.
Among various extraction parameters, pH is regarded as one of the most crucial parameter
influencing pectin yield and other physicochemical properties. In general, several acids
can be employed for pectin extraction. The more common ones include sulphuric, citric,
phosphoric, hydrochloric (HCl) and nitric (HNO3) acids.5 In this study, the latter two acids
at two pH levels were used and their effects on the pomelo pectin yield were evaluated. As
shown in Table 1, it is evident that pectin extraction at pH 2 resulted in a significantly
higher pectin yield than that at pH 3 for the extractions either by HCl or HNO3 acids. This
was possibly due to the enhanced ability of the acid in solubilising the protopectin from the
albedo with increasing acid strength. However, it is important to point out that a strong
acid solution could disadvantageously lead to smaller pectin molecules as a result of
premature hydrolysis. Comparatively, much lower pectin yield (6-14% d.b.) was reported
in the literature, where pectin was sequentially extracted using water and oxalate chelating
agent from the pomelo albedo.6 Furthermore, it is noticeable that the nitric acid, at pH 2,
was a more effective solvent for pectin solubilisation compared to its counterpart
hydrochloric acid. Such finding is also agreeable with the extraction of pectin from
buttercup squash flesh.11 From this result, it was determined that subsequent extraction of
the pectin to be done using nitric acid at pH 2 .
Typically, commercial pectins are produced from apple pomace and citrus fruits. The
former generally contains about 2-19% pectin, whereas the latter yields approximately 6-
26% pectin.4 According to the results presented herein, it was evident that the pomelo peel
View Online
yield was comparable to those from apple pomace and citrus fruits, thus, signifying the
potential use of pomelo peel as an alternative source for the commercial pectin production.
Table 1 Yield, degree of esterification and average molecular weight of pomelo pectins
extracted from different conditions.
Extraction condition
Yield (%) DE (%) MW (kDa)
Acid pH
HCl 2 19.33 0.36 64.1 0.2 645 r 5
3 11.06 2.80 70.7 0.4 562 r 7
HNO3 2 24.26 0.08 59.4 0.9 440 r 3
3 8.32 0.08 62.5 0.7 482 r 12
In order to categorise the extracted pomelo pectin according to the degree of esterification
(DE) and molecular weight (MW), samples were analysed for their DE and MW using
FTIR and gel permeation chromatography, respectively. It was observed that the pomelo
pectin spectrum possessed similarities in the absorption pattern to that of a commercial
pectin (data not shown). The DE determined using the peak areas of the free carboxyl and
esterified groups was in the range of 59-71% (Table 1). In particular, the extraction at pH
12/04/2014 12:05:06.
3 gave pectin with slightly higher DE than that at pH 2. Nevertheless, the DE of all
pomelo pectins was higher than 50%, thus, classifying them as high methoxyl pectin. This
was expected as an acidic extraction process often yields a high methoxyl pectin (DE !
50%) whereas a classical extraction through hot water typically yields a low methoxyl
pectin (DE < 50%).2
Furthermore, all extracted pomelo pectins showed relatively high MWs in the range of
440-645 kDa (Table 1). These MW values are mostly comparable to that of a commercial
high methoxyl pectin (Genu JMJ from CP Kelco), which has a MW of 480 kDa. It is
interesting to note that both the DE and MW have been found to relate to the stabilising
property of pectin. For instance, it has been shown that the stability of drinking yogurt
decreases with decreasing molecular weight of pectin.12
Temperature sweep tests were performed on various pectin dispersions, ranging from 0.4-
1% w/v, and the results are shown in terms of storage (Gc) and loss (Gcc) moduli versus
temperature upon cooling in Figure 1. The results showed that pectin dispersion was
approximately a liquid-like solution at 0.4% w/v. However, it changed from a
predominantly liquid-like solution at elevated temperature to a typically gel-like structure
with Gc > Gcc at low temperature for pectin concentrations between 0.4-0.8% w/v. The
gelling temperature (Tgel), at which Gc = Gcc, showed a tendency to increase with increasing
pectin content. At 1%w/v, pectin dispersion was more concentrated which was reflected
by the fact that Gc was exceeding Gcc across the studied temperature. For this
concentration, the gelling temperature could not be determined as it was out of the studied
temperature range of 80oC. It is worth pointing out that at above 60 oC, the pectin sample
started to dry out and its rheological response was erroneous above this temperature.
View Online
Figure 1: Temperature sweep responses of pomelo pectin dispersions at (a) 0.4, (b) 0.6,
(c) 0.8 and (d) 1% w/v.
Figure 2: Mechanical spectra of pomelo pectin dispersions at (a) 0.4, (b) 0.6, (c) 0.8 and
(d) 1% w/v.
View Online
Furthermore, the amplitude sweep experiments revealed that the moduli increased as the
pectin concentration increased from 0.4 to 1% w/v and the linear viscoelastic range was
beyond 0.1 Pa, justifying the use of such stress in the subsequent frequency sweep
measurements (data not shown). The mechanical spectra of pectin dispersions are shown
in Figure 2. It was apparent that pectin dispersion at 0.4% w/v behaved as a viscous liquid
with Gcc > Gc at low frequency and vice versa at frequency higher than about 1 Hz and the
spectrum was highly frequency dependent. At higher pectin contents, a typical weak gel
characteristic with Gc > Gcc was observed. In particular, the gel strength was expectedly
increased and its frequency-dependence diminished with increasing pectin content as a
result of higher polymer chain associations.
Viscosity curves of various pectin dispersions are shown in Figure 3. Likewise to other
biopolymer dispersions, these pectin dispersions exhibited a Newtonian behaviour at low
concentration and a shear thinning characteristic at high concentration. In this case, pectin
samples behaved as a Newtonian fluid at the concentration below 0.2 %w/v. The shear
thinning behaviour was caused by the rearrangement in the conformation of pectin
molecules in the dispersions as a result of shearing.13 Furthermore, it was evident that
increasing pectin content promoted the zero shear viscosity and shear thinning
characteristics. The zero shear viscosities (K0) were deduced and plotted as a function of
pectin concentration in Figure 4. It was noticeable that the zero shear viscosities increased
with increasing pectin content. In particular, two distinguishable regions, corresponding to
12/04/2014 12:05:06.
the dilute and concentrated regimes, could be observed. These findings are consistent with
previously reported works.14-15 At the dilute regime, the zero shear viscosities were found
to be approximately proportional to c1.8, while at the concentrated regime, K0 v c7.6, where
c is the pectin concentration. Furthermore, the critical concentration (c*) of the studied
pomelo pectin dispersions was approximately at 0.41% w/v. It is worth noting that the
rheological properties of pectin dispersions are principally dependent on pectin structure.
Previous works have proposed that an increase in the branching of pectin could result in a
higher zero shear rate viscosity, a higher shear rate dependence of viscosity and a higher
storage modulus.16-17 In particular, they also concluded that pectin side chains had little
effect on the slopes of the viscosity as a function of concentration in the dilute regime, but
higher branched pectin led to higher slopes in the concentrated regime. This conclusion
may be indicative that the pectin side chains are likely involved in the entanglements of the
pectin molecules in the concentrated regime. Further work is undergoing in order to
investigate the rheological properties of pomelo pectin gels as affected by the addition of
different divalent ions or varying acid-sugar conditions.
4 CONCLUSIONS
Pectins were extracted from pomelo peels through different acid and pH treatments. It was
shown that pH had a predominant role in controlling pectin yield. At the same pH, the
extraction using nitric acid induced higher yield compared to HCl. However, the reason
behind the difference is still unclear. All extracted pectins had a DE and MW in the range
of 59-70% and 440-650 kDa, respectively. In particular, the extraction at pH 3 appeared to
give slightly higher DE than that at pH 2. Furthermore, the gelation of pomelo pectins
showed the gelling temperature and elastic modulus increased with increasing pectin
concentration. Pectin dispersions behaved as a viscous liquid for concentrations below
View Online
0.4% w/v and as a weak gel between 0.4-0.8% w/v. Above 1% w/v, a typically strong gel,
with tanG ~ 0.2, was observed. In addition, the dispersions below 0.2 %w/v were
Newtonian but they exhibited a pseudoplastic with shear thinning characteristic at higher
concentrations.
12/04/2014 12:05:06.
References
1. J. Visser and A. G. J. Voragen, Pectin and pectinases, Elsevier, The Netherlands,
1996.
2. D. D. Joye and G. A. Luzio, Carbohydr. Polym., 2000, 43, 337-342.
3. C. D. May, Carbohydr. Polym., 1990, 12, 79-99.
4. F. Munari, M. C. Tanzi and P. Petrini, Int. J. Biol. Macromol., 2012, 51, 681-689.
5. M. H. Norziah, E. O. Fang and A. A. Karim in Gums and Stabilizer for the food
industry 4, ed. G. O. Philips, D. J. Wedlock and P. A. Williams, Oxford: IRC Press, 1999,
p 27.
6. A. Chaidedgumjorn, U. Sotanaphun, N. Kitcharoen, P. Asavapichayont, M. Satiraphan
and P. Sriamornsak, Pharm. Biol., 2009, 47, 521-526.
7. S. Piriyaprasarth and P. Sriamornsak, Carbohydr. Polym., 2011, 83, 561-568.
8. G. D. Manrique and F. M. Lajolo, Postharvest Biol. Tec., 2002, 25, 99107.
9. M. S. Lima, E. P. Paiva, S. A. C. Andrade and J. A. Paixo, Food Hydrocolloids,
2010, 24, 17.
10. S. Wang, F. Chen, J. Wu, Z. Wang, X. Liao and X. Hu, J. Food Eng., 2007, 78, 193-
200.
11. E. M. ODonoghue and S. D. Somerfield, Food Hydrocolloids, 2008, 22, 1326-1336.
12. H. A. Deckers, C. Olieman, F. M. Rombouts and W. Pilnik, Carbohydr. Polym., 2000,
6, 361-378.
13. M. A. Rao, Rheology of fluid and semisolid foods, Aspen Publishers, Gaitherberg,
MD, USA, 1999.
14. M. A. V. Axelos, J. F. Thibault and J. Lefebvre, Int. J. Biol. Macromol., 1989, 11,
12/04/2014 12:05:06.
186-191.
15. T. D. Chou and J. L. Kokini, J. Food Sci., 1987, 52, 1658-1664.
16. J. Hwang and J. L. Kokini, Carbohydr. Polym., 1992, 19, 41-50.
17. J. Hwang, Y. R. Pyun and J. L. Kokini, Food Hydrocolloids, 1993, 7, 39-53.
INFLUENCE OF STORAGE ON THE WATER BINDING OF PECTIN:
DETERMINATION BY DSC
1 INTRODUCTION
Citrus pectins, necessary for the food industry and many other applications, are produced
by large companies and distributed worldwide. On their way to the customers, the pectins
may be stored under unfavourable environmental conditions for longer periods. This can
alter their properties and affect their quality in the final application.
12/04/2014 12:05:09.
material properties such as pectin powder density or particle size and morphology proved
to be very important. These properties result from varying pectin processing conditions and
affect the availability of hydrophilic groups. For instance, small, rough and porous pectin
particles were able to adsorb water more rapidly whereas bigger particles with a smooth
surface needed some more water and also time for plasticization before close water binding
could take place.4
It can be assumed that (1) pectins from different suppliers, produced from varying raw
materials and under different processing conditions, have also varying material properties
and water binding behaviour and that (2) the storage of pectins at 60 C and 80 % humidity
might alter these properties and that, especially, the increase of hydrophilic groups might
increase also their water binding ability. Therefore the aim of the present study was to
investigate the water binding properties of different commercial pectins prior to and after
storage by DSC at different water contents. Some samples were additionally investigated
by dynamic vapour sorption in order to characterise their water uptake properties.
2.1.Materials
The water binding properties of eight commercial pectins from three suppliers (1, 2 and 3),
three high-methoxylated (HMP) and five low-methoxylated (LMP) samples, were tested.
The LMP were prepared by acidic treatment (LMP-AC), enzymatic treatment (LMP-ENZ)
12/04/2014 12:05:09.
sample type GC 0 GC S GC DM 0 DM S DM IV 0 IV S IV
% % % % % % cm3/g cm3/g %
1H HMP 85.5 93.2 109.0 59.6 51.2 85.9 598 398 66.6
2H HMP 65.8 82.7 125.6 76.9 39.6 51.5 660 348 52.7
3H HMP 80.9 87.6 108.2 69.8 43.7 62.6 554 260 46.9
1L LMP-AC 94.0 99.9 106.3 25.5 15.4 60.3 301 180 59.8
2LE LMP-ENZ 67.6 84.4 124.8 31.7 17.2 54.3 500 290 58.0
3LE LMP-ENZ 81.5 90.7 111.3 30.2 18.2 60.1 336 255 75.9
2LA LMP-AMID 61.3 68.6 112.0 29.6 21.7 73.3 450 323 71.8
3LA LMP-AMID 68.4 75.9 111.0 32.2 21.1 65.6 382 223 58.4
View Online
2.2 Methods
The DSC experiments and calculations were made as described previously,4 but using
another instrument (STA 251, NETZSCH, Germany) and a different cooling temperature.
Prior to DSC-tests, all original and stored samples were kept in a desiccator above P2O5 for
at least 7 days. The dry samples (5-10 mg) were filled into aluminium pans and wetted in a
desiccator by water sorption from the environment at aw 1 for 2, 4 or 6 h and 1, 2 or 3 d,
respectively. In order to get higher water contents, additional 5, 10 or 15 l water were
added to some of the wetted samples after three days. All pans were sealed after water
sorption or addition of water. Samples with added water were kept overnight for complete
water distribution. Additionally, all samples were measured also in the dry state.
Altogether, 10 DSC-measurements at different water contents were made for each pectin
sample.
DSC measurements were performed as a four-step procedure (cooling to -60C, heating to
80 C, second cooling to -60 C and second heating to 80 C) and melting temperatures
and enthalpies were determined. After DSC measurement, the sample pans were pierced,
dried at 120 C for 120 min, cooled down in a desiccator and weighed for determination of
the dry mass of the sample ms. From this value and the water content in the wetted sample
mw the total amount of water Wc was calculated as:
mw
Wc (g / g)
ms
12/04/2014 12:05:09.
The results from the second heating step were used for the calculations of freezing-bound
water (Wfb) with melting points below or close to 0C and free water (Wf) with melting
points clearly above 0 C. Non-freezing water (Wnf) was calculated as:
Wnf Wc (W fb W f )( g / g ) .
The dynamic vapour sorption tests (DVS) of the three HMPs were made in the Federal
Institute for Material Research and Testing using a DVS-1 (Surface Measurement Systems,
London, UK).
DSC curves of pectins differed remarkably with increasing water content Wc . Samples that
contained only none-freezing water Wnf showed no DSC peak at all (not shown). Figure 1a
gives the typical shape of freezing-bound water, a single peak with a maximum at -9 to -2
C, as detected for samples with low Wc < 0.8 g/g. Figure 1b shows a less homogeneous
peak with the maximum shifted to +1.5 to +3.5 C (indicating free water) and a shoulder (=
turning point) of freezing-bound water between 0 and +2 C. In the present study it was
typical for samples with Wc from 0.8 to 1.5 g/g (wetted for 1 to 3 d). Figure 1c was typical
for samples with Wc above 1.5 g/g. The melting peak became much broader, the maximum
shifted up to 10 C (free water) and a shoulder of freezing-bound water was often found
already at -7 to 0 C. This peak shape was found for samples with 5 to 10 l water added
after 3 d storage. In Figure 1d the water content of the sample was so high (Wc > 2.5 g/g)
View Online
that the free water peak had no clear maximum, became dominating and nearly concealed
the Wfb shoulder. This image was typical for samples with a surplus of 15 l added water.
12/04/2014 12:05:09.
Figure 1: Typical DSC curves of pectins at water content < 0.8 g/g (a), between 0.8 and
1.5 g/g (b), between 1.5 and 2.5 g/g (c) and > 2.5 g/g (d).
DSC examinations of water binding are able to detect an influence of storage on poly-
saccharides. In the case of cellulose samples, storage at high humidity (100 %) altered the
contents of non-freezing and freezing bound water because of changes in the matrix
structure of the polymer.10
The results of the DSC examination of pectins varied in dependence on the total water
content Wc. All dry samples before storage contained between 0.03 and 0.14 g/g water
(Fig. 2) and showed no peak in DSC, they contained only non-freezing water Wnf.
water (g/g)
0.1
0.05
0
1H 1L 2H 2LE 2LA 3H 3LE 3LA
sample
The water content of these samples was relatively low and depended on pectin type as well
as on the supplier. Pectins from supplier 1 contained clearly more water than those from
supplier 2 and 3. The HMPs contained less water than the LMPs from the same supplier
because of less hydrophilic groups in the molecules. That means that in dry samples
mainly the number of hydrophilic groups determines the amount of bound water and this is
in agreement with the previous results found for model pectins.4
The water binding of wetted samples will be discussed for two different moisture levels:
low water content Wc 0.8 g/g, resulting from wetting by water sorption up to three days,
and high water content Wc > 2.7 g/g, reached by adding 10 to 15 l water to the samples
after sorption. Since it is nearly impossible to get samples with identical absolute water
contents, the discussion of wetted samples is made using water fraction shares.
At low Wc the main share of water (> 80 %) was bound closely as non-freezing water. It
was expected that this share would increase after storage because of the higher number of
hydrophilic groups in the stored pectins. However, this was only found in three of eight
samples, 1 HMP and 2 LMPs (Fig. 3). In two samples Wnf even slightly decreased (1
HMP, 1 LMP). The share of freezing-bound water Wfb was much smaller than that of Wnf
(< 15 %). It clearly decreased in four of eight stored samples (2 HMP, 2 LMP), slightly in
two others and increased in the last two samples (Fig. 3). A possible explanation for this
decrease might be a reduction of size and / or number of capillaries and small inter-particle
pores in the samples by storage. The share of free water Wf was rather small at low Wc
(mostly <10 %, not shown), because of the limited total water content. Altogether, pectins
with low Wc had slightly different water binding properties before and after storage. The
differences were not systematic and independent of pectin type or supplier.
12/04/2014 12:05:09.
water (%)
100 15
90
80 10
70
60 5
50
40 0
1H 1L 2H 2LE 2LA 3H 3LE 3LA 1H 1L 2H 2LE 2LA 3H 3LE 3LA
sample sample
Figure 3: Non-freezing and freezing-bound water, associated with pectins at low water
content (about 0.8 g/g), in original and stored samples.
At high Wc > 2.7 g/g the relation of the different types of water in DSC measurements
changed considerably. The share of non-freezing and freezing-bound water decreased and
the share of free water increased up to 45 % (Fig. 4). The differences between the
individual pectins with respect to the water types were higher than at low Wc.
The effect of pectin storage on water binding at high Wc was also not systematic. There
was an increase in Wnf in five samples and a decrease in three others. The Wfb and Wf
decreased in three and four samples, respectively. Once more, pectin type and supplier had
no systematic influence on the alterations.
View Online
Wnf
water (%)
water (%)
original stored
20
original stored Wfb
100
90 15
80
70 10
60 05
50
40 00
1H 1L 2H 2LE 2LA 3H 3LE 3LA 1H 1L 2H 2LE 2LA 3H 3LE 3LA
sample sample
water (%)
50
original stored Wf
40
30
20
10
00
1H 1L 2H 2LE 2LA 3H 3LE 3LA
sample
Figure 4: Non-freezing, freezing-bound and free water associated with pectins at high
water content (>2,7g/g) in original and stored samples.
12/04/2014 12:05:09.
An interesting difference was found in the shape of DSC curves of original and stored
pectin samples with a medium water content around 1.5 g/g. DSC signals of original
pectins had a clear Wnf shoulder slightly below or around 0 C and a single melting peak of
free water (sometimes with a small Wf shoulder at the end) with a maximum between 1
and 6 C. The corresponding stored samples had a less clear shoulder of Wfb. The
maximum of the free-water peak increased and a much more pronounced shoulder on the
right side of the peak maximum or even an additional free water peak above 7 C was
formed. Typical examples are the DSC curves of HMP 2H (Fig. 6) and LMP-ENZ 2LE
(Fig. 7). One explanation was that stored pectins were not able to bind as much water as
the originals and that the surplus water formed small ice crystals of free water on the lid or
wall of the pan, which broadened the melting peak of stored samples.
Figure 5: DSC melting curves of pectin 2H with Wc 1.8 g/g; a=original and b=stored
pectin.
View Online
Figure 6: DSC melting curves of pectin 2LE with Wc 1.5 g/g; a=original and b=stored
pectin. c d
Another possibility might be that stored pectins were able to trap more water in a gel-like
layer than the originals and that this water had a different melting range than the typical
free water. In general, high shares of freezing-bound and free water might form very broad
overlaying peaks or even double-peaks and the clear determination of the fractions is
difficult.
Results of DVS measurements support the different effect of the storage on the sorption
behaviour of individual pectins. In case of HMP 2H the isotherms of the stored sample are
rather close to or slightly below those of the original whereas in the case of HMP 3H the
stored sample isotherms are slightly above those of the originals (Fig. 5).
12/04/2014 12:05:09.
Figure 7: Sorption and desorption isotherms of HMP 2H (a) and 3H (b); full lines =
original pectins, dotted lines = stored pectins; full symbols = sorption, open symbols =
desorption.
CONCLUSIONS
The presented results of the water binding of commercial pectins are only partly in
agreement with those found for the model pectins prepared in laboratory scale.4 For both
pectin groups the water content of dry samples was determined as expected by the number
of hydrophilic groups (HMP < LMP) and varied in the case of the commercial pectins in
View Online
storage it is necessary to examine the material properties more intensively by methods such
as BET-surface area measurement or mercury intrusion porosimetry. This is in preparation
and will be published later on.
References
1. INTRODUCTION
The color on the colored rice is from the color of rice bran which has been classified into
seven color classes: white, light-brown, speckeled brown, brown, red, variable purple, and
purple by the USDA National Small Grains Collection (NSGC). Most people eat colored
12/04/2014 12:05:11.
rice as the whole instead of polished rice because it is believed that dark color of bran
containing substances with great health benefits for human beings. Recently, the
consumption of whole grain cereal has been strongly recommended by many governmental
and health organizations. Many epidemiological studies have indicated that consumption
of whole grain cereal is highly correlated to reduced incidences of chronic diseases.1 The
bioactive phytochemicals, such as phenolic compounds that are rich in the whole grains,
may be a mechanism whereby whole grains have their protective effects.2 Most
phytochemicals in the whole grain rice are present in the bran fraction consisting of bran
layers (pericarp, seed coat, nucellus, and aleurone) and the germ. The majority of phenolic
compounds in bran are bound covalently to cell wall components.3 However, only the free
phenolic compounds in the digestive tract can take action, either on site and/or at remote
sites after absorption, against the incidence of colon cancer and other chronic diseases.
Rice bran is one of the rich sources of dietary fiber and contains high concentrations of
arabinoxylans (AX).4 The physiological effect of cereal bran arabinoxylans and modified
AX has brought great attention in both academia and industry.5 In addition, the
arabinoxylan oligosaccharides (AXOS) degraded from AX by physical or enzymatic
treatments have been thought to provide bifidogenic effects and immune activity6 that can
be used in functional foods.
Ultrafine milling or grinding is an important unit operation in many fields such as
material industries and pharmacy.7 The powders with reduced particle sizes will increase
their surface area which creates several important properties or functions, including the
increase of hydrophilic property, content of soluble dietary fiber etc.8-10 In this study,
ultrafine milling was applied to reduce the size of colored rice bran to improve the
solubility and bioaccessability of rice bran that could be used alone or combined with other
ingredients for neutroceutical products.
Usually, the whole grain products or brown rice have less acceptance than the refined
or polished rice by consumers, because the products are less palatable or tasty with fibers
View Online
from bran. To grind whole grain or bran by ultrafine milling is an efficient way to reduce
the roughness and granular mouthfeel of products because very fine particles or powder,
usually less than 25 Pm, can be obtained. In addition to improve the mouthfeel of products
containing rice bran with very fine particles, the increase of surface area of particles might
increase the bioaccessibility of bioactive compounds in rice bran.11 Most fibers in rice bran
are water insoluble. However, water soluble fibers or oligosaccharides have better
biofunctionality for chronic diseases prevention. Therefore, reducing the particle size of
rice brain by physical treatment, such as ultrafine milling applied in this study, or
degrading the AX by endohydrolysis enzymes might increase the amount of water
extractable AX (WE-AX) for AXOS preparation. Two major objectives of this study were
(1) to evaluate the effects of dry and wet ultrafine ball-milling on the increase of the
amount of WE-AX, and (2) to prepare AXOS from IDF of colored rice bran by combining
ultrafine ball-milling and enzymatic hydrolysis.
2.1 Materials
In our preliminary studies, 6 colored rice (three black rice and three red rice) collected
from local markets were investigated. However, the appearance of endosperm after
polishing and their starch characteristics based on the amylose content were inconsistent
among the samples and within each sample. Therefore, only the highest purity of colored
12/04/2014 12:05:11.
rice, Hualien Taibalang black waxy rice (HB) and Hualien Taibalang red waxy rice (HR),
purchased from Hualien KuangFeng Farmers Association (Hualien, Taiwan), were used in
this study.
The colored rice was polished using a laboratory rice mill (VP-31T, Yamamoto Co., Ltd.,
Tendu, Japan) with the setting of 5 for flow and the whiteness. The colored rice was
polished for 2-4 passes until the milled rice was close to white. Rice bran collected from
the first pass and from the second to fourth passes of milling were defined as outer-layer
rice bran (ORB) and inner-layer rice bran (IRB), respectively. The bran samples were
stored at 4C until analysed.
The rice bran was immediately stabilized by tempering 16 g rice bran to 20% moisture
content and being heated in a microwave (900 w) for 30-180 s. The content of free fatty
acid (FAA) in treated bran was monitored during two months of storage at 4 and 25.
Both dry and wet ultrafine ball-milling (DBM and WBM) were carried out on the
stabilized rice bran (SRB) to reduce the particle size of ORB. A planetary ball mill (PM
100, Rersch, Haan, Germany) with 50 ZrO2 balls (10 mm in diameter) in a 250 mL ZrO2
milling ball was used to reduce the particle size and increase the solubility of rice bran by
milling at the speed of 400 rpm for various times (0, 24, 48, 72 and 120 hr).
For AXOS preparation, insoluble dietary fiber (IDF) fractionated from the defatted
rice bran, as called RBIDF, which was prepared by removing most of the protein and
View Online
starch with protease, amylase and glucoamylase from defatted ORB were used as the
starting materials.
Both the sugar profile and the amount of WE-AX and the Arabinose/Xylose (A/X)
ratio in ultrafine ball-milled SRB and RBIDF were determined.
The amount of WE-AX were determined by completely hydrolyzing both stabilized rice
bran sample (SRB) and IDF isolated from defatted rice bran (RBIDF) and their extracts
into simple sugars. The sugars were determined by using a high-performance anion-
exchange chromatography equipped with a pulsed amperometry detector (Dionex ICS-
3000, Dionex Co., Sunnyvale, CA, USA). The AX content and A/X ratio were calculated
by the equations of (Ara% + Xyl%) 0.88 and Ara%/Xyl%, respectively.
Sugar composition of non-cellulosic carbohydrates in colored rice bran samples (SRB
and RBIDF) were also analyzed using HPLC-PAD.
Both RBIDF and WBM-RBIDF was treated with Pentopan, an endoxylanase from
Thermomyces lanuginosus (2500 U/g, Sigma-Aldrich Inc., MO, USA) to prepare AXOS
before and after ultrafine wet ball-milling. The degree of hydrolysis was determined by the
increase of reducing sugar ratio.
12/04/2014 12:05:11.
The molecular size distribution of water extractable material was determined by using
high-performance size exclusion chromatography (HPSEC) equipped with a RI detector
(L-3300, Hitachi Ltd., Tokyo, Japan). To monitor the molecular size distribution soluble,
the TSK G4000PWXL and G2500PWXL columns (TOSOH, Tokyo, Japan) were used. The
eluent was 50 mM NaNO3 with 0.02% NaN3 and the flow rate of mobile phase was 0.5
ml/min. The system was calibrated with standards, pullulans (MW = 2.12105, 1.12105,
4.73 104, 2.28 104, 1.18 104, and 5.9 103), xylooligosaccharides (xylohexose,
xylopentaose, xylotetraose, xylotriose, and xylobiose), glucose, and xylose.
Table 1 shows the starch properties and the degree of milling of two colored rice samples
(HB and HR) used in this study. Both colored rice were waxy rice with amylose content
less than 2%. The black rice (HB) had higher amount of outer layer rice bran with the
similar chemical compositions to the ORB from red rice (HR).
View Online
Based on the free fatty acid contents of microwave-heating treated ORB, a good storage
quality of rice bran at 4 and room temperature could be obtained by microwave-heating
the rice bran for 120 sec (Figure 1). If the roast flavor is preferred for rice bran applying
12/04/2014 12:05:11.
for the ingredients in a premix or foods, longer heating time, such as 150 or 180 sec
microwave-heating could be applied.
Table 2 shows sugar composition, content of arabinoxylan and A/X ratio of the SRB and
RBIDF. High A/X ratio of SRB and RBIDF indicated that the xyloses in the AX backbone
were highly substituted with arabinose in colored rice bran compared to other cereal bran.
After removing most of starch from rice bran, the content of AX in RBIDF was almost
doubled and the A/X ratio was slightly increased.
Figure 1 Effect of microwave-heating treatments (0 (control), 30, 60, 90, 120, and 150 sec)
on the free fatty acid content of rice bran (HR) during storage at 4 (a) and 25
(b).
View Online
Table 2 Sugar composition, content of arabinoxylan and A/X ratio of stabilized colored
rice bran (SRB) and IDF fractionated from the defatted rice bran (RBIDF)
HB HR
Sugar SRB RBIDF SRB RBIDF
(%, db) (%, db)
Arabinose 1.81 0.06 2.04 0.00 2.36 0.04 4.10 0.89
Xylose 1.76 0.11 1.84 0.05 2.10 0.04 3.16 0.64
AX1 3.14 0.14 6.42 0.05 3.92 0.07 6.38 0.15
Ara/Xyl 1.03 0.04 1.11 0.03 1.12 0.01 1.30 0.01
Fucose 0.58 0.04 0.52 0.07 0.29 0.06 ND2
Rhamnose 0.06 0.03 0.09 0.02 0.06 0.01 ND
Galactose 1.03 0.04 1.10 0.08 0.98 0.03 1.60 0.13
Glucose 40.83 1.13 47.65 3.11 25.09 1.05 1.94 0.05
1
AX = 0.88 (% arabinose + % xylose), with the factor 0.88 to correct for water uptake in the
hydrolysis procedure.
2
ND means not detectable.
Figure 2 shows the color differences of ball-milled HB and HR rice bran. The picture (a)
12/04/2014 12:05:11.
and (g) were the original coarsely milled rice bran and the pictures (b) to (e) and (h) to (k)
were the ultrafine dry-milled black rice bran and red rice bran, respectively. It is found that
the color fated after ball-milling and the discolor was obvious when milling time was
prolonged. Pictures (f) and (l) were the frozen-dried wet-milled samples, no significant
discolor was observed because of lower friction heat produced during wet-milling
compared to dry-milling.
According to the results of particle size distribution, both dry- and wet ball-milling could
effectively reduce the particle size of rice bran and produce a narrower particle size
distribution (Table 3). In fact, the wet ball-milling was more efficient than the dry ball
milling in a shorter milling time. The amount of WE-AX significantly increased in both 24
Figure 2 Appearance of dry ball-milled HB bran (a-e) and HR bran (g-k) for 0, 24, 48, 72
and 120 h, respectively; liphilized wet ball-milled (8 hr) HB (f) and HR (l) bran.
View Online
Table 3 Particle size, WE-AX% and A/X ratio of HB-SRB before (BM0) and after dry
(DBM) and wet (WBM) ball-milling for 24 and 8 hr, respectively
BM0
Character DBM24 W-BM8
(control)
Particle size
Mean 93.75 8.42 4.75
d10 4.18 2.19 1.09
d50 62.86 6.93 2.85
d90 223.10 17.21 12.27
Characteristics of WE-AX
% of total AX 6.76 38.78 34.3
A/X 1.96 1.13 1.0
hr dry- milled and 8 hr wet-milled rice bran, those were 6 and 5 times increases,
respectively. The A/X ratio of WE-AX changes after milling, which a lower A/X ratio was
found in the milled samples.
During dry-milling, the particle size slightly increased instead of decreasing when
extending the milling time up to 120 hr (Figure 3(a)). This is attributed to the aggregation
of warm particles, resulting in less efficiency on size reduction than the wet ball-milling
(Figure 3(b)).
12/04/2014 12:05:11.
Figure 3 Changes of particle size distributions of dry (a) and wet (b) ball-milled HB-SRB
View Online
Figure 4 Effects of dry (a) and wet (b) ball-milling on the WE-AX% (: HB & :HR)
and their A/X ratios (: HB & : HR) in stabilized colored rice bran.
In order to compare the particle size on the efficiency of enzymatic reaction on the AX, the
particle size distributions of both dry and wet ball-milled RBIDF were determined. It is
12/04/2014 12:05:11.
found that after wet ball-milling for 16 hr the smallest particles could be obtained (Figure 5)
so that the WBM-RBIDF was followed by the treatment of Pentopan, an endoxylanase, for
AXOS preparation.
From the size-exclusion chromatogram (Figure 6), it was found that only a small amount
of water soluble carbohydrates with lower molecular weight existed in the IDF without
ball-milling pretreatment. While, the major water soluble carbohydrates with 36.4 kDa and
some oligosaccharides could be found in the 16 hr wet-milled IDF.
Figure 5 Particle size distributions of HR bran IDF before (BM0, control) and after dry
ball-milling (DBM) for 24, 48, and 72 hr and wet ball-milling (WBM) for 16 hr.
View Online
Figure 6 Molecular size distribution of water soluble mzterial from RBIDF (HB) before
(BM0) and after wet ball-milling for16 hr (WBM16). Standards (from left to
right): pullulans (MW=2.12105, 1.12105, 4.73104, 2.28104, 1.18104, and
5.9 103), xylooligosaccharides (xylohexose, xylopentaose, xylotetraose,
xylotriose, and xylobiose), glucose, and xylose, respectively.
After Pentopan treatment, reducing sugar contents in both IDF and WBM-RBIDF
increased and almost double amount of reducing sugar could be obtained in IDF with ball-
milled pretreatment (Figure 7). These results indicated that the combination of wet ball-
milling and enzymatic hydrolysis could be an effective way to prepared the AXOS from
RBIDF.
5. ACKNOWLEDGEMENTS
This work was kindly supported by the grants from the Council of Agriculture (98-2815-C-
002-158-B and 99AS-3.1.4-FD-Z2), Executive Yuan, Taipei, Taiwan.
References
Madhav P. Yadav1*, Fei Zhang, Tu Luan2, Lijiao Wu2, and Hongbin Zhang2
1
Eastern Regional Research Center, Agricultural Research Service, U. S. Department of
Agriculture, 600 E. Mermaid Lane, Wyndmoor, PA 19038, USA.
E-mail: madhav.yadav@ars.usda.gov
2
Advanced Rheology Institute, Department of Polymer Science and Engineering, Shanghai
Jiao Tong University, Shanghai 200240, China
1. INTRODUCTION
Corn fiber (CF), the most abundant low-valued by-product of the industrial corn wet-milling
process, contains a high percentage of valuable arabinoxylan (AX), which can be isolated by
alkaline hydrogen peroxide as corn fiber gum (CFG) (1, 2). CFG consists of a -1, 4 linked
D-xylopyranosyl backbone with -L-arabinofuranosyl substituents attached at positions 2
and/or 3 and most of glucuronic acid at position 2 with the following glycosyl composition: D-
xylose (4854%), L-arabinose (3335%), galactose (711%), and glucuronic acid (36%) (3,
12/04/2014 12:05:14.
4, 5, 6, 7, 8, 9). It shows a good emulsification ability for oil-in-water emulsions which may be
due to presence of functional components (protein, phenolic acids and lipid) covalently
attached to the carbohydrate polymer chains (10).
Recently a great interest has been developed in formulating products by blending a
mixture of gums to produce a desired product, which can combine the advantage of each gum
component and also bring some additional excellent properties (11, 12). For instance, a
mixture of locust bean gum and carrageenan makes a more viscous solution than the sum of
their individual viscosities (13). The rheological properties of these mixed systems are
essentially important for both scientific and industrial aspects (14). For example, the
stabilization of emulsions in food or cosmetics depends considerably on the viscosity of the
mixed system (15).
Considering an excellent emulsifying ability and low solution viscosity of CFG, it is
important to study its viscosifying action when mixed with an aqueous solution of different
polysaccharides or their derivatives to broaden its applications for many food and non-food
uses. Thus the objective of the present work is to study the thickening properties of CFG when
it is mixed with either anionic, cationic or neutral carbohydrate polymer and propose a
possible model of interaction in their aqueous solution.
Mention of trade names or commercial products in this publication is solely for the purpose of providing
specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture.
USDA is an equal opportunity provider and employer.
View Online
2. EXPERIMENTAL METHODS
2.1. Solution preparation. The polysaccharides were dispersed in distilled water and mixed on
a roller mixer for 24 hours. The dispersion of methylcellulose (MC) was transferred to a
refrigerator at 4 oC until completely solubilized as it did not dissolve well at room temperature.
The chitosan (CTS) solution was prepared in 1% acetic acid. For the mixture, the CFG
solution at each concentration was mixed with polysaccharide solution at room temperature
and mixed gently on a roller mixer.
2.2. Rheological measurements. The rheological measurements were carried out using a
rotational rheometer AR G2 (TA Instruments, USA) with a 218 cone plate geometry (60 mm
diameter and 58 m gap). The temperature was regulated by a circulating water bath using a
peltier system. A thin layer of low-viscosity silicone oil was placed on the surface of the
solution held between the plates to reduce the evaporation of water from the samples during the
measurement. Steady shear viscosity was measured over a shear rate range of 0.01-1000 s-1 at
25 oC. Steady shear state was assumed to be attained, when the variation of torque was less
than 5% throughout three consecutive sampling periods (20 s). The maximum point time was
set as 6 minutes.
The viscosity behavior of the mixture of CFG and hyaluronan (HA) was investigated and the
viscosifying action of CFG on HA solution was evaluated in detail. Figures 1 and 2 show the
effect of shear rate on the steady shear viscosity and the steady shear stresses of individual
12/04/2014 12:05:14.
CFG, HA and their mixtures at 25 oC. The steady shear viscosity of CFG solution at different
concentrations was almost independent of the shear rate even at a very high shear rate of up to
1000 s-1, showing a Newtonian fluid behaviour. The steady shear viscosity of CFG solution
increased with increase in its concentration but still it was as low as 0.3 Pa s even at a
relatively high concentration (60 mg/mL). The viscosifying action of CFG in its solution with
HA (10 mg/ml) became more prominent as its concentration increased from 10 to 60 mg/ml.
The zero shear viscosity of individual HA (10mg/ml), CFG (10, 30 and 60 mg/ml) and HA
mixtures are shown in Figure 3. It was very clear that the zero shear viscosity of the HA/CFG
mixture containing 60 mg/ml CFG increased more than five times (above 35 from about 7 Pa
s) in comparison to the HA/CFG mixture containing only 10 mg/ml CFG, though the viscosity
of individual CFG (60mg/ml) solution was only about 0.3 Pa s.
The ability of CFG to form low viscosity solutions is a good indication of its highly
branched structure as reported for other highly branched synthetic (16) and natural polymers
(17). A pseudoplastic fluid behaviour was found for HA/CFG mixture, which was similar to
the individual HA solution, indicating its rheological properties are dominated by HA. Also
the viscosity of HA/CFG mixture was much higher than the algebraic sum of its individual
components showing a significant viscous synergism. The steady shear viscosity curve of
polyamidoamine, PAMAM (a typical dendrimer) is shown in Figure 4, which indicates a very
clear Newtonian behavior with very low viscosity even at its high concentration (60mg/ml). A
reduction in the viscosity of HA solution was seen on the addition of dendrimer showing its
antagonistic effect unlike the synergistic effect of CFG as shown in Figure 1. Figures 5 and 6
View Online
show the steady shear viscosity of individual CFG and its 1:1 mixture with chitosan and MC
respectively at 25 oC.
The solution properties of biopolymer mixtures can be quite different than its each
component. Such effect often leads to an increase in viscosity of mixed polymers solution in
comparison to individual component at the same concentration (18). The quantification of thje
viscosity change in the mixed polymers is always advantageous for industries, as their
products are usually formulated with more than one additive to achieve the desired physical
structure and properties (19). To characterize the viscosifying action of CFG, a shear stress
synergism index, I S , as defined by the following expression (20) was used:
W i j
Is (1)
W i W j
where W i j , W i , and W j are the mean shear stresses in the whole range for shear rate for i + j, i
and j at the concentrations Ci + Cj, Ci and Cj respectively. The mean values of shear stress were
calculated from the rheograms W f (J ) by the following expression:
1 J2
W
J1 J2 1
J
WdJ (2)
where J1 and J2 are the minimum and maximum shear rates to which the aqueous system are
subjected.
According to this definition, when I S >1, the shear stress of the mixed system would be
12/04/2014 12:05:14.
larger than the algebraic sum of its component viscosities, i.e., synergism would result. The
magnitude of I S reflects the degree of synergism that varies with composition of the mixed
polysaccharide solution. When I S <1, it is regarded as antagonism.
The shear stress synergism index, ( I S ), as calculated using equation (1) for the solution
of two individual components ( W i , and W j ) and their mixed solution W i j , with various weight
fractions of CFG under the whole shear rate range are given in Table 1. Apparently, the
synergistic effect of CFG was found in all studied HA/CFG mixtures and the I S value
increased from 1.263 to 2.102 with increasing CFG concentration from 10 to 60mg/mL
solution. A similar viscosifying action of CFG was also observed in its mixed solution with all
three kinds of polysaccharides (cationic, anionic and neutral). The synergistic effect of CFG
does not look closely related to the ionic or non-ionic feature and chemical structure of the
added polysaccharides, since when CFG is mixed with an anionic polymer (HA), a cationic
polymer (chitosan) and a neutral polymer (MC), it showed a similar steady shear viscosity
with a slight difference in its degree of viscosifying action. The CFG also showed a very
remarkable synergistic effect in its mixture with the following additional polysaccharides:
guar gum, hydroxyethyl cellulose (HEC), konjac glucomannan, pectin, carboxymethly
cellulose (CMC) and chitosan with a shear stress synergistic index ( I S ) values 1.24, 1.22,
1.26, 1.80, 1.28 and 1.45 respectively (Table 1).
The synergistic interaction model for CFG and all the above mentioned non-gelling
polysaccharides is more likely similar to the one proposed for CFG/HA (Figure 7). It seems
more likely that the electrostatic and steric repulsions between CFG and HA resulting from the
View Online
highly branched structure of CFG are not the barriers. The hydrophobicity cannot be
considered as the main driving force of interaction for such highly hydrophilic
macromolecules. So the hydrogen bonding between CFG and the non-gelling polysaccharides
used in our studies has been proposed as the main driving force for the viscous synergism.
Competition for solvent and differential swelling, which alters the effective concentration in
the mixtures, may also contribute to the viscosifying action of CFG.
12/04/2014 12:05:14.
Figure 1: Steady shear viscosity of individual CFG, HA solution and their mixtures at
25 C.
View Online
Figure: 2. Shear stress as a function of shear rate of individual CFG, HA solution and
12/04/2014 12:05:14.
their mixture at 25 C.
Figure 3: Zero shear viscosity of individual CFG, HA solution and their mixture at 25 C
View Online
Figure 4: Steady shear viscosity of individual HA, PAMAM solution and their mixture at
25 C.
12/04/2014 12:05:14.
Figure 5: Steady shear viscosity of individual CFG, chitosan solution and their mixture at
25 C.
View Online
Figure 6: Steady shear viscosity of individual CFG, MC solution and their mixture at
12/04/2014 12:05:14.
25 C
Table 1: The shear stress and corresponding shear stress synergism index (IS) of individual
polymer solution and their mixture with CFG under the whole shear rate range (25 C).
4. CONCLUSION
12/04/2014 12:05:14.
An aqueous solution of CFG shows a Newtonian fluid behavior. The viscosity of CFG
increases very slightly with increase in its concentration even at a very high shear rate
The solution of CFG shows synergistic behavior with all three kinds (anionic, cationic and
non-ionic) of polysaccharides. The degree of synergism varies with the composition of the
mixed polysaccharide and shear rate. The viscosifying effect of CFG on the solution of
other polysaccharides is due to their intermolecular interaction, more likely by hydrogen
bonding with OH groups present on their smooth regions. The viscous synergism may also
be caused by the segregation of other polysaccharide due to its effective concentration
increase after addition of CFG in the same solution. The property of making a low
viscosity solution and the synergistic effect of CFG on other carbohydrate polymers can be
of a great commercial importance for formulating many food and non-food products.
Acknowledgement
The authors are pleased to acknowledge Dr. Kevin B. Hicks for his proper guidance in
many steps and National Natural Science Foundation of China for providing financial
support (Grant No. 21274090).
View Online
References
15. Y. Hemar, M. Tamehana, P. A. Munro, & H. Singh, Food Hydrocolloids, 2001, 15 (46),
565-574.
16. I. Sendijarevic & A. J. McHugh, Macromolecules, 2000, 33 (2), 590-596.
17. X. Li, Y. Fang, S. Al-Assaf, G. O. Phillips, K. Nishinari & H. Zhang, Food Hydrocolloids,
2009, 23 (8), 2394-2402.
18. I. Donati, I. J. Haug, T. Scarpa, M. Borgogna, K. I. Draget, G. Skjak-Braek, & S. Paoletti,
Biomacromolecules,2007, 8 (3), 957-962.
19. J. Pellicer., J. Delegido, J. Dolz, M. Dolz, M. J. Hernandez, & M. Herraez, Food Science
and Technology International, 2000, 6 (5), 415-423.
20. M. Dolz, M. J. Hernndez, J. Pellicer, & J. Delegido, Journal of Pharmaceutical Sciences,
1995, 84 (6), 728-732.
NON-LINEAR DYNAMIC VISCOELASTICITY OF XANTHAN GUM SOLUTIONS
1 INTRODUCTION
Xanthan gum is a high molecular-weight polysaccharide, soluble in cold water, and mainly
used in the food industry as a thickener and stabiliser. Therefore, the rheological properties
of xanthan gum are of key interest to improve their thickening and stabilising properties.
Traditionally, the rheological properties of xanthan gum have been determined by flow
12/04/2014 12:05:16.
curves and small amplitude oscillatory shear (SAOS). It is shown that xanthan gum
possesses a high viscosity at low shear rates, which is related to its strong stabilising
properties, whereas its shear thinning behaviour allows the system containing xanthan gum
to flow when a shear force is exerted1. Recently, large amplitude oscillatory shear (LAOS)
measurements have been carried out in order to gain a deeper insight into microstructural
changes in complex fluids such as xanthan gum solutions2, 3. Furthermore, LAOS results
may be useful to describe the elastic and viscous properties of complex fluids at large
deformations, which are closer to real processing conditions.
In this work we have studied the influence of salt concentration on the rheological
properties of xanthan gum solutions, comparing the results obtained by means of SAOS
and LAOS. We have analysed the non-linear oscillatory response on the framework
proposed by Ewoldt et al.4 in order to obtain parameters with meaningful physical
meaning. Briefly, they defined two kinds of parameters to determine the non-linearities
present at different steady state cycles (intracycle non-linearities):
x Elastic Chebyshev coefficient, e3, and viscous Chebyshev coefficient, v3. Material
behaviour is classified in four categories: strain-softening (e3 < 0), strain-hardening
(e3 > 0), shear thinning (v3 < 0) and shear thickening (v3 > 0).
S
G '
L GM'
(1)
GL'
K L' K M'
T (2)
K L'
12/04/2014 12:05:16.
2.1 Materials
2.2 Methods
2.2.1 Solution Preparation. The preparation method employed was similar to the one
used in the food industry. First, the gum was slowly added to Milli-Q water with
mechanical stirring (Ikavisc MR-D1) at room temperature for 3 hours. Afterwards, the
solution was heated to 70C for 45 min under continuous stirring. Then, the sodium
chloride was added to the heated gum solution. In order to remove the air bubbles within
the solution, it was sonicated for at least 1 hour. The solutions were stored at 4C for 24-48
hours before they were measured.
gap = 0.053 mm. All the tests were performed at 20 C, using a solvent trap to inhibit
evaporation.
Strain sweep tests were carried out in the strain range from 1 to 300% at a fixed frequency
of 1 rad/s. Frequency sweep tests were conducted in the linear region from 20 to 0.05 rad/s.
The LAOS analysis required raw strain and the stress signal which were acquired by means
of native control software (TA Orchestrator) using the Arbitrary wave shape test as
described by Ewoldt et al5.
Data processing
The raw data obtained from the arbitrary wave shape test was processed with the
MITLAOS software5. This software was used to calculate the Fourier coefficients,
Chebyshev coefficients, decomposition of stress and the viscoelastic moduli such as, GL,
GM, KL and KM.
View Online
First, we have explored the influence of salt concentration in the range 0-0.5 wt% NaCl for
0.4 wt% xanthan gum solution by means of SAOS measurements. Figure 2 shows the
storage (G) and loss (G) moduli as a function of strain. Although it is reported that
addition of a small amount of salt modifies the viscosity of xanthan gum solutions1,6,
Figure 2 shows no effect of salt addition on G and only a small deviation of G in the
non-linear viscoelastic region. Furthermore, the same critical strain value (29% 3%)
which defines the linear viscoelastic region (LVR) was obtained for all cases.
Figure 3 shows G* and loss tangent (tan G as a function of frequency within LVR
(strain = 5%). The results are not affected by salt concentration as shown by G*, only a
slight deviation of tan G values at high frequencies for the xanthan gum solution without
NaCl was observed. This behaviour is due to the fact that the absence of salt provokes an
increase of G at the higher frequencies covered, while G is unaffected. The slight
increase of the viscous response at higher frequencies could indicate some microstructural
changes for the system without salt. Analysing the mechanical spectra (data not shown) it
was observed that the frequency dependence of the elastic modulus can be quantitatively
described by a power law. A value of 0.25 was obtained for the slope of all the G vs. Z
curves in the 0.05 2 rad/s range. The solutions studied exhibited weak-gel viscoelastic
behaviour as demonstrated by this slope value and by the fact that G values lay above
those of G. These results are consistent with those previously reported for xanthan gum
12/04/2014 12:05:16.
solutions7-9.
Figure 2: Storage (G) and loss (G) moduli as a function of strain for xanthan gum
solutions (0.4 wt%) with different contents of NaCl: 0 wt%, 0.025 wt%, 0.1 wt% and 0.5
wt% at 20C and at 1 rad/s. Sensor system cone-plate Diameter=50 mm, angle=0.04 rad,
gap=0.053 mm.
View Online
Figure 3: Frequency sweeps (20-0,05 rad/s) of xanthan gum solutions (0.4 wt%) with
different contents of NaCl: 0 wt% (S,U), 0.025 wt% (,), 0.1 wt% (,) and 0.5
(z,{) wt% at 20C. G* (Pa) and tan (G) vs frequency (rad/s). Sensor system cone-plate
(Diameter=50 mm, angle=0.04 rad, gap=0.053 mm).
12/04/2014 12:05:16.
Figures 4a and 4b show the first harmonic (average) elastic modulus and first harmonic
(average) dynamic viscosity, respectively for the 0.4 wt% xanthan solutions with NaCl
ranging from 0 to 0.5 wt% at 1 rads-1. The elastic modulus decreases beyond the critical
strain (30%) indicating a strain softening behaviour. The elastic response is not clearly
affected by salt concentration as it was also shown in SAOS results. Furthermore, a slight
growth of the dynamic viscosity of the solutions was observed when increasing NaCl
concentration. This could be due to the shielding of the electrostatic repulsion between the
charged side chains of xanthan gum which could lead to more compact structures. This
explanation is also supported by the fact that the solution without NaCl is the one which
presents a small strain overshoot. This feature is related to an increase of dynamic viscosity
at the onset of the non-linear region before the decrease of the viscosity, and it is
associated with a structural rearrangement preceding the collapse of the highly extended
structure of xanthan gum in the absence of salt10, 11.
A deeper insight into the influence of the salt concentration on the microstructure of
xanthan gum can be obtained by analysing the non-linear response. In order to obtain
meaningful physical information the experimental LAOS results have been analysed on the
basis of the framework proposed by Ewoldt et al.4
Figures 4c and 4e show the parameters used to evaluate the intracycle elastic non-
linearities; third order elastic Chebyshev coefficient ratio, e3/e1; and strain-stiffening ratio,
S. e3/e1 ~ 0 and S ~ 0 are obtained within the LVR. At strains above the critical one, two
behaviours were observed. At strains closer to the LVR both parameters were slightly
negative, which corresponds to a strain-softening behaviour. However, at larger strains the
View Online
sign of the parameters change from negative to positive indicating a transition from strain-
softening to strain-stiffening behaviour. Furthermore, a clear difference at large strains was
observed between the xanthan gum solutions without salt and with salt (not depending on
the amount of salt in the range studied 0.025 wt% 0.5 wt%). This should be related to the
more extended polymer chains of the xanthan gum solution in the absence of salt which
may offer more resistance to the elastic deformation. It is worth noting that a significant
difference between systems with and without salt was detected by LAOS, which was not
previously shown by SAOS tests.
Figures 4d and 4f show the parameters used to evaluate the intracycle viscous
nonlinearities; the third order viscous Chebyshev ratio v3/v1; and the shear-thickening ratio,
T. Again, at strains within LVR the values of v3/v1 and T are 0. At strain of 100% and
300% the sign of the parameters are positive, whereas at the highest strain 1000% the sign
becomes negative. The positive value of these parameters indicates a shear-thickening
behaviour. This feature should be related to the well-known weak strain overshoot
response obtained for xanthan gum solutions3,7,10. This behaviour was slightly detected by
SAOS tests only for the system without salt. However, LAOS revealed the occurrence of a
weak strain overshoot for all the studied systems. The existence of this peak is due to a
structural rearrangement as explained above. The shear-thickening behaviour is more
noticeable when no salt is added to the solution, which can be associated with a higher
resistance to the flow accounting for the higher number of entanglements of the extended
polymer in the absence of salt. At higher strains the intracycle behaviour became strain-
thinning, being this character more pronounced for that without salt. This could be
12/04/2014 12:05:16.
explained taking into account that once the structure is disrupted the more extended
xanthan gum chains may be aligned easier with flow due to its lower molecular rigidity.
4 CONCLUSIONS
Figure 4: Oscillatory shear tests of xanthan gum at 1 rad/s analysed by means of LAOS
parameters (0.4 wt% aqueous with 0 wt%, 0.025 wt%, 0.1 wt% and 0.5 wt% NaCl. a) First
harmonic (average) elastic modulus G'1. b) First harmonic (average) dynamic viscosity
K'1. A red line has been added as guide for the eye to highlight the occurrence of weak
strain overshoot c) e3/e1. d) v3/v1. e) Strain-stiffening ratio, S. f) Shear-thickening ratio, T.
View Online
ACKNOWLEDGMENTS
The financial support received (Project CTQ2011-27371) from the Spanish Ministerio de
Economa y Competitividad (MINECO) and from the European Commission (FEDER
Programme) is kindly acknowledged.
References
1. G. Sworn, in Food Stabilisers, Thickeners and Gelling Agents, 1st Edn., ed. A. Imeson,
Wiley-Blackwell,Oxford, 2010, ch. 17, p. 325-342.
2. R. H. Ewoldt, P. Winter, J. Maxey, G. H. McKinley, Rheol. Acta, 2010, 49, 191.
3. K. Hyun, M. Wilhelm, C. O. Klein, K. S. Cho, J. G. Nam, K. H. Ahn, S. J. Lee, R. H.
Ewoldt, G. H. McKinley, Progress in Polymer Science, 2011, 36, 1697.
4. R. H. Ewoldt, A. E. Hosoi, G. H. McKinley, J. Rheol., 2008, 52, 1427.
5. R.H. Ewoldt, C. Clasen, A.E. Hosoi, G.H. McKinley, Soft Matter., 2007, 3, 634
6. K. S. Kang, D. J. Pettit, in Industrial Gums, 3rd Edn., ed. R. L. Whistler, J. N.
BeMiller, Academic Press, San Diego, 1993, ch. 13, p. 354-357.
7. K.-W. Song, H.-Y. Kuk, G.-S. Chan, Korea-Australia Rheology Journal, 2006, 18, 67.
8. R. Pal, AIChE J., 1995, 41, 783.
9. M. M. Talukdar, I. Vinckier, P. Moldenaers, R. Kinget, J. Pharm. Sci., 1996, 85, 537.
10. K. Hyun, S. H. Kim, K. H. Ahn, S. J. Lee, J. Non-Newtonian Fluid Mech., 2002, 107
51.
11. F. Lequeux, P. Hebraud, J. P.Munch, D. Pine, Proceedings of The Second World
12/04/2014 12:05:16.
Department of Food and Nutritional Sciences, University College Cork, Cork, Ireland
Oxidised cellulose and oxidised starch are entirely different products. The reagent
normally used in manufacture of oxidised starch is sodium hypochlorite which (like
periodate) cleaves the C(2)C(3) bond of glucose, leaving aldehyde groups at both
positions; on further oxidation these are converted to carboxyl groups. The reaction is
12/04/2014 12:05:18.
H H H H H H
COO COO OH
H H OH H H
O O HO
HO
O O O
O
HO HO O
H H O H H
OH OH COO
CH2OH
H H H H H H
Slope = 2.0
1 G'
0 1
log (G"/Pa)
G"
-1 5%
0.1
-2 K*
-3 0.01
0.4 0.6 0.8 1.0 1.2 0.1 1 10 100
log (c/wt %) Frequency (rad/s)
Figure 2: (a) Concentration-dependence of G' (z) and G" (O), measured at 1 rad/s and
1% strain, for dispersions of MDOC in water. (b) Frequency-dependence of G' (z),
G" (O) and K* (S) for 5.0 wt % MDOC. All measurements were made at 20C.
(Figure 2a), and at higher values of concentration (c) the slope of log G' versus log c
approaches the limiting value of 2 (c2-dependence) commonly observed for gelling
biopolymers.6
At 5 wt % (Figure 2b) and above, the dispersions give mechanical spectra5 with
the form typical of a gel network:7 G' is about an order of magnitude higher than G",
both show only slight variation with frequency (Z), and complex dynamic viscosity (K*)
has a slope close to -1 when plotted double-logarithmically against frequency.
Gel-like character was also evident5 from creeprecovery experiments in which strain
was monitored over a 5 min period of applied stress and for a further 5 min after the stress
was removed. Figure 3 shows the results obtained for 5 wt % MDOC at a stress of 0.9 Pa.
0.5
0.3
Strain
0.2
5 % MDOC
0.1 0.9 Pa
0
0 1 2 3 4 5 6 7 8 9 10
Time (min)
When the stress is applied, there is a sharp initial increase in strain, as would be seen for
deformation of an elastic solid. This is followed by a slow, progressive increase, with the
variation of strain versus time becoming linear, as would be seen for a viscous liquid.
When the stress is removed, there is a sharp reduction in strain, demonstrating survival of
an elastic network with solid-like properties. At the end of the recovery period, however,
there is still a residual ("irrecoverable") strain, which can be attributed to re-arrangement of
network structure during the creep period (i.e. corresponding to the progressive increase in
strain in response to the applied stress).
At higher values of applied stress, above ~ 1 Pa, the network fractured, and response
was then dominated by viscous flow.5 Materials that show predominantly elastic, gel-like,
response to small deformations but break down and flow like liquids at higher stress are
known as "weak gels".7 Since close-packing of spherical particles does not occur until the
volume-fraction reaches ~65 %, development of "weak gel" properties (Figures 2 and 3) at
a concentration as low as 5 wt % indicates that the individual particles of MDOC associate
with one another to form a crosslinked network, rather than interacting solely by physical
contacts.
The purpose of the research described in the following section was to explore the
effect of guar gum, chosen as a typical disordered polysaccharide with extensive practical
applications as an inexpensive thickener, on the self-association and "weak gel" rheology
of MDOC.
Six samples of guar gum were used: Meyprogat M7, M30, M60, M90 and M150 from
Meyhall and a standard food-grade sample from Sigma. The intrinsic viscosities of these
materials and the values of molecular weight derived from them by the Mark-Houwink
relationship reported by Picout & Ross-Murphy8 are listed in Table 1.
Dispersions of MDOC in distilled deionised water were prepared at a fixed
concentration of 10 wt % by overhead stirring for 20 min at ambient temperature.
Solutions of guar gum, also in distilled deionised water, were prepared at twice the
required final concentrations, and were mixed with an equal weight of the MDOC
dispersion, giving an MDOC concentration of 5 wt % in all samples. The mixtures
were then stirred for 30 min, using a magnetic stirrer.
Low-amplitude oscillatory measurements of G', G" K* were made at a fixed strain
of 1 %, using cone-and-plate geometry. After loading, samples were coated around their
periphery with light silicone oil, to minimise evaporation, and held for an ageing period
of 2 h. A mechanical spectrum was then recorded. Temperature was held fixed at 20C.
Table 1: Intrinsic viscosity ([K]) and molecular weight (MW) of guar gum samples
_____________________________________________________________________
Sample M7 M30 M60 M90 M150 Sigma
_____________________________________________________________________
[K] (dl g-1) 1.30 3.85 6.11 9.55 14.25 12.10
MW (kDa) 60 280 538 1013 1786 1420
_____________________________________________________________________
View Online
2.2 Results
All mixtures gave mechanical spectra with the same general features as the spectrum
shown in Figure 2b for 5 wt % MDOC with no added guar gum (G' > G"; little variation
in moduli with frequency; linear reduction in log K* with increasing log Z).
Figure 4 shows observed changes in G' (at 1 rad/s) with increasing concentration of guar
gum in the mixtures with 5 wt % MDOC for the Meyprogat samples of lowest and highest
molecular weight (M6 and M150) and for one of intermediate molecular weight (M90). In
all cases there is a steep initial increase in G' with increasing concentration of guar gum.
The values then pass through a sharp maximum, and decrease again as the concentration of
guar gum is raised further. The initial increase in modulus can be explained by increased
self-association of MDOC in response to the presence of guar gum, and the subsequent
decrease by excessive association, leading to formation of large aggregates that make little
contribution to crosslinking of the "weak gel" network.
Initial enhancement and subsequent loss of gel strength as the extent of intermolecular
association is increased is a common feature of biopolymer gelation. For example, the
moduli of gellan gels9 rise to a maximum and then decreases again on progressive addition
of either monovalent or divalent metal ions or on progressive acidification. Kappa carrageenan
similarly passes through a maximum in gel strength with increasing concentration of
calcium10 or alkali metal11 cations, and maxima in gel strength on varying salt concentration
and/or pH are commonly observed in thermogelation of globular proteins.12 Perhaps the best
known example is Ca2+- induced gelation of alginate,13 where cation concentrations beyond
the value required for maximum gel strength cause precipitation.
12/04/2014 12:05:18.
Indeed, initial increase and subsequent decrease in gel strength with increasing degree
of self-association is to be expected, and has been described14 as a "Goldilocks effect".
In the classic story of Goldilocks and the Three Bears, the little girl liked her porridge to be
"not too hot, not too cold, but just right" In the same way, maximum gel strength will occur
at an optimum degree of crosslinking: less association will give a weaker network; greater
association will give larger aggregates, with consequent reduction in the effective number of
individual junctions, until ultimately the network collapses into a solid precipitate.
M150
4 M90
3 M7
G' (Pa)
0
0.001 0.01 0.1 1
[Guar gum] (wt %)
Figure 4: Variation of G' (1 rad/s) with concentration of guar gum in mixtures with
5 wt % MDOC, illustrated for Meyprogat samples M7 (O), M90 (S) and M150 (z),
with molecular weights (Table 1) of, respectively, ~60, ~1000 and ~1800 kDa.
View Online
0.1
Cmax (wt %)
0.01
0.001
0 500 1000 1500 2000
MW (kD)
Figure 5: Effect of molecular weight on the concentration of guar gum needed to give
maximum enhancement of G' in mixtures with 5 wt % MDOC for Meyprogat samples (z)
and the sample of molecular weight ~1420 kDa from Sigma (O).
Although all three traces in Figure 4 show a similar initial increase and subsequent
decrease in G' with increasing concentration of guar gum, the values of concentration
at which maximum enhancement occurred (cmax) are grossly different, ranging from
12/04/2014 12:05:18.
References
1 O.B. Wurzburg, Modified Starches: Properties and Uses, CRC Press, Boca Raton,
Florida, 1986.
2 C. Bertocchi, P. Konowicz, S. Signore, F. Zanetti, A. Flaibani and S. Paoletti,
Carbohydrate Polymers, 1995, 27, 295-297.
3 S. Wie, V. Kumar and G. Banker, International Journal of Pharmaceutics, 1996,
142, 175-181.
4 V. Kumar and T. Yang, Carbohydrate Polymers, 2002, 48, 403-412.
5 A.A. Agoub and E.R. Morris, Carbohydrate Polymers, 2008, 71, 416-427.
6 A.H. Clark and S.B. Ross-Murphy, British Polymer Journal, 1985, 17, 164-168.
7 S. B. Ross-Murphy in Biophysical Methods in Food Research, ed. H. W.-S. Chan,
Critical Reports on Applied Chemistry, SCI, London, 1984, pp.195-290.
8 D.R. Picout and S.B. Ross-Murphy, Carbohydrate Research, 2002, 337, 1781-1784.
9 E.R. Morris, K. Nishinari and M. Rinaudo, Food Hydrocolloids, 2012, 28, 373-411.
10 J. Doyle, P. Giannouli, K. Philp and E.R. Morris in Gums and Stabilisers for the
Food Industry 11, eds. G.O. Phillips, P.A. Williams and D.J. Wedlock, Royal Society
12/04/2014 12:05:18.
Joop de Vries
CSM Bakery Supplies Europe, Innovation Centre Sweet Ingredients, Fruitlaan 24, NL-
4462 EP Goes, the Netherlands
1 INTRODUCTION
Glazing gels are being used in bakeries to give shine and protection (from drying and
discolouring) to fruit on open fruit pies, to cakes with a 'bavarois' top layer, and to baked
goods like croissants. In most cases, they are applied by spraying (typically at 85C) with a
spray gun (e.g. Bakon Food Equipment) from a bag-in-box packaging or containers (up to
1000kg), but it is possible to apply them simply by brushing. The gel should than set
immediately after depositing. However, over-heating by keeping the gel at high
temperatures for too long times, or under-heating, can result in under-performing. But also
for convenience reasons, as well as to protect the fruit from heat damage, bakers would
appreciate a cold-prepared glazing gel. This can be achieved by mixing two phases (e.g. a
12/04/2014 12:05:19.
Ca-source and an alginate or pectin), but that usually results in too fast gel formation
leaving no time to apply it on the cake. Other options: a viscous liquid, or a liquid with a
yield stress, but liquids do not have a 'dry touch', and stick to the plastic foil used to
separate different cakes in a box. Applying a liquid that forms a gel with calcium from the
fruit1 is yet another option, but the diffusion of Ca usually is too slow to prevent run-off
before gel setting. The best option so far is to use a gel with thixotropic, pseudoplastic or
shear-thinning properties. This paper compares weak gels with two gelling agents: low-
methoxyl amidated pectin Genupectin LM104AS from CPKelco (Degree of
Esterification=26, Degree of Amidation=22, no Ca added to the gel) and Na-alginate (high
M type, Grindsted FD155 from DuPont Danisco). Screening of weak gels of a wide
variety of gelling agents learned that these two gelling agents performed best (easy to
'liquefy' without leaving gel lumps by stirring with a wire whisk or spoon, and formation of
a 'dry-touch' gel within an hour).
'Thixotropic behaviour' (in the practical use of this word) was assessed in a rheometer test
(oscillation, followed by shear, and oscillation again) as shown in Figure 1. Gels were
made by dissolving pectin and alginate in water with 0,1% Na-citrate buffer, adding sugar,
heat to 90, add acid and only in the case of alginate - a small amount of Ca-citrate (5%
of stoechiometric Ca-saturation of the alginate), cool to 40C before packaging and storage
at ambient temperature for at least one week. The gels (1% gelling agent) had 30% sugar,
and a pH of 3,6. Comparison of the two gels shows that the alginate gel recovers very fast,
whereas the LMA-pectin gel needs a few minutes for recovery. Despite this difference,
View Online
both products fulfill the requirements for successful application on fruit pies: directly after
application, the gels have sufficient consistency not to flow off the fruit, and within an
hour the gel has set.
12/04/2014 12:05:19.
Figure 1 Recovery from shear of a weak LMA-pectin gel (bottom) and a weak alginate gel
(top). Oscillation with plate-plate geometry for 4 minutes, followed by shear at shear rate
50/s, after which oscillation re-starts.
View Online
Figure 2: Heat-sensitivity of a weak alginate gel (top) in comparison with a weak LMA-
pectin gel (bottom). Oscillation in cup-and-bulb geometry at increasing temperature.
It is clear from Figure 2 that the LMA-pectin melts at around 50C, whereas the alginate
gel as expected- does not melt below 95C. In most cases, heat-resistant hydrocolloid gels
have a high gel-setting temperature, but for weak alginate gels it is not possible to measure
a distinct gel setting temperature. Unexpectedly, weak alginate gels do not set on cooling
(Figure 3). Cooling directly after preparation from boiling temperature to room
temperature or refrigerator temperature does not induce gelling, but the gel is formed when
stored at low temperatures for longer time (days). It is known that the binding of Ca to
View Online
alginate is fast and strong even at elevated temperatures. Therefore, we assume that gelled
areas are formed on addition of Ca to the hot alginate solution, and that there is a slow
transfer of Ca from these gelled areas to the dissolved Na-alginate, resulting in a gel after
several hours /days. The gel formed in this way is very heat resistant (Figure 2). It is clear
from these results that weak LMA-pectin gels and weak alginate gels have a different
12/04/2014 12:05:19.
gelling mechanism (time and temperature play different roles in the two cases), but both
gels can recover from shear. For use as cold-prepared fruit glazing gels, it is desirable to
have a product that has the same properties at room temperature and refrigerator
temperature: storage of the glazing gel product as well as storage of the fruit can take place
at both temperatures (fresh fruit is normally stored cool, unlike canned fruit). The ratio of
the gel strength (both measured as breaking strength by penetration, and with low-
deformation rheology as storage modulus G') at 20C to the gel strength at 10C is about
0.75 for weak alginate gels, but much lower for weak LMA-pectin gels (about 0.35),
therefore weak alginate gels outperform weak LMA-pectin gels in this application. The
LMA-pectin gels discussed so far did not contain any Ca, therefore it is interesting to
compare their behavior in the presence of Ca.
The gelling mechanism of pectin includes several types of bonds: OH hydrogen bonds,
NH2 hydrogen bonds, CH3 hydrofobic bonds, Na / K/ Ca binding, and specific Ca-binding
in 'egg-boxes'. It is generally assumed that specific Ca-binding plays an important role in
the gelling mechanism of LMA-pectin. However, most commercial LMA-pectin have a
DFA (degree of free acids) of about 50% (i.e. 50% of the galacturonic acid residues is not
esterified or amidated). Assuming a random distribution of ester groups, statistics learn that
the percentage of galacturonic acids present in blocks of 5 or more is 0,55 = 0.03, in other
words for a molecule with MW=80.000 (Degree of Polymerisation = 400), there are only
12 galacturonic acid residues available per molecule, perhaps just sufficient to provide the
12/04/2014 12:05:19.
two blocks per molecule that are needed as a minimum for network formation. And indeed,
LMA-pectins are advised for use in thermo-reversible gels, whereas Ca-binding of the
'egg-box' type is known to be heat irreversible. Comparing the thermal behavior of a weak
conventional LM-pectin gel with a weak LMA-pectin gel of equal gel strength both at
their optimal Ca dosage- confirms that commercial LMA-pectins with DFA of around 50%
do not form 'egg-box'-like Ca-binding2: the conventional LM-pectin gels do not melt
(crossover of storage modulus G' and loss modulus G'' in oscillatory viscosimetry),
whereas the LMA-pectin gel melts at temperature below boiling point. Increasing the gel
strength of LMA-pectin gels can make them heat-irreversible (the melting temperature
increases with the gel strength), but decreasing the gel strength of LMC-pectin gels with
sufficient Ca cannot make them fully heat-reversible. LMA-pectins with a higher DFA
than 50%, or LMA-pectins made to contain galacturonic acid blocks (e.g. by pectin methyl
esterase), do have the possibility to form 'egg-boxes' with Ca. For the application in fruit
glazing gels, weak LM-pectin gels with Ca have the disadvantage that it is impossible to
make a gel without some 'structure' visible as a result of the very fast reaction of Ca with
conventional LM-pectins.
Chapter 2.1 concluded that alginate weak gels outperform LMA-pectin weak gels as cold-
prepared fruit glazing gels because of the lower temperature-sensitivity of their gel
strengths, so the temperature of use at the bakeries is less critical. Another requirement is
shelf stability: the gels should have a shelf life of at least 6 months at ambient temperature.
Refrigeration of these type of ingredients (glazes) is not common in the bakery industry.
Figure 4 shows that the shelf life of alginate weak gels at ambient temperature is not
sufficient:
View Online
140
130
rel.gel strength G' (day 2=100%)
120
110
LMA-pectin 4C
100
LMA-pectin 20C
90
alginate 4C
80
alginate 20C
70
60
50
40
0 5 10 15 20 25 30 35 40 45 50
storage tim e (days)
Figure 4: Shelf stability of weak LMA-pectin and weak alginate gels (pH=3,6).
Alginate solutions are known to show viscosity decrease on ambient storage, but alginate
gels are ambient stable. Figure 4 shows that weak alginate gels are not stable at ambient
temperature. This can be due to chemical degradation like acid hydrolysis or oxidative
12/04/2014 12:05:19.
degradation, but the decrease of the gel strength can also be due to changes in the gel
build-up. Most gels increase their gel strength for a period of weeks after gel preparation,
like our weak gels do at refrigerator temperature. Alginate and pectin are both composed of
uronic acids, but their stability to degradation (acid hydrolysis, -elimination) is far from
comparable as is shown in Figure 5. It is impossible to find a pH-range where both pectin
and alginate is stable against degradation at elevated temperatures: a problem in cases
where a combination of these two gelling agents is being used. At the acidic range of fruit
glazing gels (pH=3,5 4), alginate is very instable at higher temperatures, and it is
therefore likely that the instability at room temperature is due to the same mechanism: acid
hydrolysis. So, unfortunately, the alginate-based fruit glazing gel has insufficient shelf
stability at the low pH that is desired because of taste and of microbiological stability. The
decreasing gel strengths are due to acid hydrolysis. Strong alginate gels are stable at mildly
acidic pH, but in alginate solutions and these weak alginate gels, slow degradation takes
place at room temperatures. The different pH-sensitivity of pectin and alginate illustrates
the different mechanisms of degradation: -elimination is important for pectin, but
surprisingly alginate is much more sensitive to acid hydrolysis compared to pectin.
View Online
120
relative viscosity (110% = directly
100
after preparation)
80
alginate residual viscosity
60
LMA pectin residual viscosity
40
20
0
3,5 4 4,5 5 5,5 6 6,5
pH
3 CONCLUSION
12/04/2014 12:05:19.
Weak LMA-pectin gels and weak alginate gels show different gel setting mechanisms: fast
setting for pectin, and extremely slow setting for alginate gels. They also show different
stabilities at ambient temperatures: weak pectin gels are stable, but weak alginate gels are
instable. The ideal 'thixotropic' gel suitable for cold-prepared fruit pie glazing gels is not
yet found. Weak alginate gels perform well, because their gel strength is less varying with
the temperature of use, but they are not shelf stable because of sensitivity to acid
degradation. Weak LMA-pectin gels only perform well when the temperature of use is kept
constant (either room temperature or refrigerator temperature, both for the glazing gel
product and the fruit pies).
References
1 O. Chevalier, I.Naudts, and J.-L. Soyeur, WO05077195 A1, August 25th, 2005.
2 P. Grant, E. R. Morris, D.A. Rees, P.J.C. Smith and D. Tom, FEBS letters, 1973, 32, 195
RHEOLOGICAL EFFECTS OF DIFFERENT INTERACTIONS IN KAPPA-
CARRAGEENAN/LOCUST BEAN GUM/KONJAC GLUCOMAMMAN GELS
Graduate School of Human Life Science, Osaka City University, 3-3-138 Sugimoto,
Sumiyoshi, Osaka 558-8585, Japan
1 INTRODUCTION
The use of plant galactomannans, mainly LBG, to modify the rheology of KC gels has
attracted attention from researchers for at least the last 40 years. KGM, derived from tubers
12/04/2014 12:05:21.
of different species of the Amorphallus konjac plant, has not featured in the English
literature on physico-chemical properties of its mixtures with KC for quite as long as plant
galactomannans, although the plant has been cultivated and powder obtained by
pulverization of ground konjac tubers has been used in different foods in East and South
East Asia for at least several centuries1. Many findings support the presence of the same
type of interaction between gluco- or galactomannans and KC. Initial2 (and more recent3)
investigations of the interaction of different galactomannans with KC involved optical
rotation measurements, and supported direct binding of mannan backbone fractions free of
galactose residues to KC in the helical conformation. Later investigations by Morris and
co-workers on crystalline fibres failed to recognize any crystalline structure that could not
be attributed to the KC helix in mixtures with either KGM4 or LBG (e.g.5), and concluded
that the binding was either not present, or did not affect the crystalline packing of KC
helices. Electron spin resonance investigations of binary KC mixtures with KGM or LBG
indicated immobilization of a fraction of the former concomitant with gelation6, 7, while the
latter was unaffected8. While these findings indicate a stronger interaction of KC with
KGM than LBG, it is not likely that direct binding is involved with the former but not with
the latter, because LBG was only partly released into the supernatant following cold
mixing and centrifugation of mixtures with KC9, and also because both polysaccharides
affected the binding of 133Cs+ ions to KC helices10.
In terms of the rheology, both KGM3, 6, 11 and LBG3, 8, 12-15 are known to increase the
elastic modulus of KC, induce KC gelation13, produce a synergy-type peak at a certain
ratio under conditions of fixed total polysaccharide content (LBG3), and increase the
fracture strain and stress in compression (LBG15, 16) and extension (KGM11). This paper
mainly summarizes previous rheological investigations on binary and ternary
LBG/KGM/KC gels17-19, and expounds on different aspects of the rheological modification.
View Online
3 RESULTS
Figure 1A shows the effect of LBG or KGM addition to a fixed (0.8 wt%) KC
concentration, while Figure 1B shows the effect of KGM addition on KC (0.6 wt%,
100mM added K+). The storage Youngs modulus increases in these binary gels up to a
plateau value. Turquois et al.15 added LBG to a lower (0.4%) KC concentration and found
that a plateau of the Youngs modulus was reached, and also saw an eventual slight
decrease, possibly due to plasticising effects. We could not dissolve enough LBG or KGM
to check for this effect. The increase in the fracture strain in extension seems to owe only
to LBG or KGM addition beyond the content leading to saturation of the elastic modulus
(i.e., reaching the plateau value), although a slight fracture strain increase is found at low
KGM additions for the K+-added sample. The weight ratio of KC to KGM or LBG at the
elastic modulus saturation will be referred to as the rheological stoichiometric ratio. When
no K+ was added, the rheological stoichiometric ratios seemed to be independent of the KC
12/04/2014 12:05:21.
concentration (and small changes in [K+]). Figure 1 demonstrates the presence of two types
of KGM or LBG binding to KC: the initial binding leads to elastically active bonds, while
binding beyond elastic saturation affects the large deformation rheology. The rheological
stoichiometric ratios are 1:5.5 (LBG:KC) and 1:7 (KGM:KC). We note that the molecular
weight of KGM is about 10% higher than that of LBG, and the polysaccharide content of
the KGM powder is also about 10% lower than that of the LBG powder (manufacturers
data). Thus, in terms of both weight and molarity, LBG can form more elastically active
bonds than KGM with KC. We will adopt the view that LBG and KGM adsorb to KC helix
crystallites (aggregated junction zones) with possible bridging of junction zones7, 9, 20, 21. It
is possible that there is a minimum distance below which helix-aggregates cannot be
bridged because both LBG and KGM are stiff molecules, so that the conformations taken
by loops and protractions away from the crystallites are restricted. The lower number of
possibilities of bridging by KGM may be related to its longer sections of unsubstituted
(and unbranched) mannan-glucan backbone1 compared with the relatively short
unsubstituted sections (rarely over 10 mannopyranose residues22) of LBG. The same
rationale could also explain why KGM-KC bonds are stronger (cf. stronger E' increase in
Figure 1A). Once E' of LBG/KC or KGM/KC gels had stagnated, it could be further
increased by addition of KGM or LBG, respectively. In the case of LBG/KC gels, the
added KGM can form stronger elastic bonds, and KC-KGM interaction is prioritised over
LBG-KGM interaction. The higher number of hydrogen-bonds between KGM and KC,
which is reflected in higher rigidity (and therefore E' increase) should also be manifested in
a higher thermodynamic compatibility between KGM and KC than LBG and KC. In the
case of KGM/KC gels, it seems that LBG can still form elastically active bonds that KGM
cannot, consistent with the lower rheological stoichiometric ratio of KGM.
View Online
Figure 1: Left, storage Youngs modulus (circles) and extension fracture strain (squares)
as a function of LBG (filled symbols) and KGM (open symbols) added to 0.8% KC.
Triangles represent data for CTot = CLBG + CKGM with CLBG () or CKGM () fixed to
0.16%. Right, E' and fracture strain in compression and extension of 0.6% KC as a
function of added KGM in the presence of 100mM added KCl.
Figure 2 shows the Youngs modulus and fracture stress and strain of ternary gels
obtained at a fixed total polysaccharide concentration. There is a synergy-like peak for
12/04/2014 12:05:21.
both binary gels at ratios close to those inferred from the data of Figure 1, 1:5.5 LBG:KC
and 1:7 KGM:KC. We note that the decrease in E' with further substitution of KC with
KGM is stronger than with LBG, probably because the importance of the total number of
bonds increases at the expense of the importance of bond strength at low KC
concentrations18. The rather flat global maximum of E' appears at a ratio of about 50:7:1
KC:KGM:LBG. The peak position is consistent with the data of Figure 1, that is, LBG can
increase the elastic modulus of KGM saturated gels. We note that as a result, addition of
1:1 LBG:KGM leads to higher E' than addition of either polysaccharide alone at KC
concentrations well below the rheological stoichiometric ratio.
The endotherms of KGM/KC, LBG/KC and LBG/KGM/KC gels were given elsewhere18.
Figure 3 gives the exotherms of these mixtures. The main features of each binary mixture
are virtually identical during cooling and heating, and are similar to those reported in
several studies3, 7, 8, 13. During heating or cooling, KC shows a single endothermic or
exothermic peak. With addition of KGM, a second peak at a higher temperature develops,
while for LBG, the temperature difference between peaks is smaller and the second peak
appears as a shoulder. Once a sufficient amount of KGM (or LBG) has been added, only
the higher temperature peak is present during cooling or heating. The higher temperature
peak corresponds to formation/melting of mixed zones of KC helices bonded to KGM or
View Online
Figure 2: Storage Youngs modulus (F = 3Hz, top) and extension fracture strain (middle)
and stress (bottom) of ternary gels (total polysaccharide content 1.2%, KC = CKC /1.2%,
KGM = CKGM /(CKGM + CLBG)). Left, results for fixed [K+], right, [K+] not fixed. Reproduced
with modification from Brenner et al.18 with permission from Elsevier.
View Online
Figure 3: Differential scanning calorimetry exotherms obtained for binary and ternary
gels containing 1.2% total polysaccharide and fixed [K+]. The ratios of the
polysaccharides of each mixture are indicated above its corresponding curve.
LBG well beyond a ratio of 1:5.5 (LBG:KC), that is, the rheological stoichiometric ratio.
We note that Kohyama et al.23 assigned the two DSC endothermic peaks during heating of
KC/KGM mixtures in the wrong order, possibly because [K+] and [Ca2+] were not fixed
and the single peak at low KC content appeared at lower temperatures.
The DSC data presented indicate that the stoichiometric ratio reflecting total amount
of bound KGM or LBG is much higher than the rheological stoichiometric ratios. We will
refer to the KGM (or LBG) chains that form elastically active bonds as population 1 and
those that still affect the DSC (i.e., are bound) but do not form elastic bonds as population
2. Population 2 clearly increases the fracture strain although it does not increase the elastic
modulus. As the data also indicate, excess chains that do not bind to KC at all also increase
the fracture strain, and can be referred to as population 3.
The evolution of the higher temperature peak follows closely the ratio of KGM to KC
in the mixtures as it has a larger effect than LBG on both the formation and melting of the
mixed network. However, the higher ratio of the higher temperature peak area to that the
lower temperature peak at the same KGM:KC ratios when LBG is present shows that in
terms of total binding, LBG and KGM bindings are additive.
Interest in the potential of the ternary gels as dessert jellies precipitated investigations in
the presence of 25% sucrose and citric acid (pH 3.4-3.7). We changed CKGM while fixing
the total polysaccharide content to 1.2%, and fixing CKC = CLBG = (1.2%-CKGM)/2. The
main findings were a decrease in E' with increasing konjac content17, a strong decrease in
E' with increasing degree of KC degradation due to heating at low pH17, and the lack of
effect of initial degradation on the fracture strain19 (Figure 4). Interestingly, in the lack of
chain degradation, a clear peak in the fracture stress arises at an intermediate KGM
addition. Note that CKGM = 0 or 0.6% correspond to [KC KGM] = [0.5 0] and [0.25 0.67],
View Online
respectively. Following the straight line between these points in Figure 2, a peak in the
fracture stress is seen in the case of fixed [K+] but not when [K+] decreased with increasing
KGM. Why, therefore, is a peak present for the sucrose-added acidic gels, where [K+] was
not fixed, despite the values of the fracture strain being closer to those of gels with lower
[K+]? Another puzzling observation is that the fracture strain is not affected by moderate
12/04/2014 12:05:21.
chain degradation. A speculation that focused on the relative weights of populations 1 and
populations 2&3 has been offered19, but as population 1 is very small compared to the total
added KGM or LBG, it is perhaps better to offer an explanation based on the effectiveness
of populations 2&3 to increase the fracture strain and stress.
4 DISCUSSION
From a practical viewpoint, because KGM affects the small and large deformation
rheology of KC (above 0.5% KC) more than LBG, KGM would be the KC-rheology
modifier of choice in the absence of contrary (financial, textural) considerations. Our
investigations of ternary gels have identified a global maximum of E' at a constant total
polysaccharide content with a KC fraction of 80-85% at additions of 7:1 KGM:LBG.
Binding has been speculated to involve individual gluco- or galactomannan chains and
an aggregate of KC helices. The main indications of mannan adsorption to the surface of
helix aggregates rather than individual helices are the lack of interaction with non-
aggregated KC helices3, the much weaker interaction with KC under conditions that
promote extensive helix aggregation7, 24 or with agarose, which forms much larger
aggregates than KC3. In addition, cold mixing of different galactomannans with KC leads
to much higher ratios of KC to galactomannan at saturation than those obtained after
heating, where large KC helix aggregates dissolve and allow binding to smaller aggregates
on subsequent cooling9. Finally, the literature offers no reports of interaction between
KGM or galactomannans and -carrageenan. It therefore appears that there is a minimum
size of helix aggregates of red seaweed galactans to which KGM and LBG chains can bind.
However, extensive aggregation beyond this minimum size decreases the binding capacity
because the total surface area of helices decreases. Conditions that promote helix
aggregation will therefore first increase, and then decrease, the total amount of KGM or
View Online
LBG that can bind to red seaweed galactans. Population 1 KGM or LBG chains stabilize
junction zones and probably form bridges between aptly positioned junction zones.
Population 2 chains are expected to form redundant bridges between junction zones, as
they do not form new connections to the elastic network, but can be sacrificed without
breaking the network. As sucrose increases the ordering of water, it is expected to promote
hydrogen bonding between LBG or KGM and KC, and affect the mixed gels more than
pure KC25. Indeed, the Youngs modulus of acidic gels with added sucrose was higher than
that of the pH-neutral gels, which we ascertained to be an effect of the added sucrose,
rather than the lower pH (unpublished data). The initial addition of sucrose to KC leads to
a higher number of smaller junction zones26, which should increase the number of
possibilities for population 2 chains to increase the fracture strain. Apparently this offsets
the inherent higher connectivity of the KC network in the absence of galacto- or
glucomannan, as the resulting fracture strains are very similar to those in the absence of
sucrose. The reason for the peak in fracture stress in Figure 4 is therefore the stronger
decrease in E' with KGM addition. The higher dependence of E' on KC concentration
reflects higher homogeneity. Addition of K+ leads to more extensive KC aggregation and
bigger junction zones, and therefore fewer possibilities for redundant bridging are present.
This explains the lower increase in fracture strain at higher [K+] (Figures 1&2). As chain
degradation should lead to smaller junction zones, excess KGM and LBG could form more
redundant bridges and the fracture strain of dessert-like gels was unaffected (or even
slightly increased) by initial stages of degradation (Figure 4).
One open question is why population 3 chains increase the fracture strain further.
Clearly, the origin of this effect is not entanglement of free chains with the network, as the
12/04/2014 12:05:21.
5 SUMMARY
References
1 INTRODUCTION
Polysaccharides such as guar gum, locust bean gum, xanthan gum or pectin are often used
in acid milk gels to modify the rheological properties.1,2,3 However it has been shown that
the addition of guar gum to skim milk at the natural pH of the mixture results in a
segregative phase separation between the casein micelles and guar gum which is described
12/04/2014 12:05:23.
Skim milk with 46 g.kg-1 of proteins was prepared with low-heat skim milk powder (CH
low heat, Ingredia, Arras, France) dissolved in Milli-Q water (made by reverse osmosis
followed by filtration through a Milli-Q apparatus) using a magnetic stirrer for 1 h at room
temperature and kept overnight in a refrigerator in order to let proteins fully hydrate. The
milk was then heat-treated at 80C during 7 min using a thermostatically-controlled water
bath. It was cooled down to 60C and guar gum (Viscogum, Cargill, USA) was dispersed
under stirring. The mixture was stirred 30 min at 60C and finally cooled to 43C. Guar
gum concentrations from 0 to 0.5 wt% were studied.
Milk was acidified by addition of different levels (1, 1.5 or 2.5 wt%) of GDL (Sigma
Chemical Co., St Louis, MO, USA) at 43C under stirring during 1 min. Immediately after
the GDL addition, the milk was poured into plastic cylinders (30 mm diameter, 65 mm
height) and placed in an incubator at 43C until a pH of 4.60 (r0.05) for 1.5 and 2.5 wt%
GDL and pH of 4.7 (r0.05) for 1 wt% GDL, respectively. The pH was measured in the
meantime with a pH-meter Consort D130 multiparameter analyzer (Turnhout, Belgium).
After cooling down to 20C during 1 h at room temperature, the gels were sheared using of
a peristaltic pump; the product passed through two successive plastic pipes (length: 40 cm
and internal diameter: 7 mm; length: 100 cm and internal diameter: 3 mm) with a mesh of
500 m holes at the end while pushed by a Masterflex peristaltic pump (Cole Parmer,
Vernon Hills-USA) (380 mL.min-1). Stirred acid milk gels were stored at 4C during 1 day
12/04/2014 12:05:23.
before analysis.
In order to speed up the macroscopic phase separation, some mixtures were
transferred prior to gelation in centrifuge tubes (25 mL) and centrifuged in a 3.18 K
centrifuge (Sigma, Munich, Germany) with a centrifuge force of 2.5 g at a temperature of
43C until pH 4.6 was reached.
For phase separation observations, a non-acidified guar gum / skim milk mixture was
poured in glass tubes (diameter 27.5 mm). Backscattering values at 880 nm were acquired
every 40 m along a 40 mm height of tube using a Turbiscan LAB (Formulaction, France)
at different time intervals.
Systems were observed using a Leica TCS AOBS SP2 (Leica, Germany) confocal
laser scanning microscope (CLSM). Proteins were labelled by addition of DyLight 549
(20 L in 1 g of sample) that was excited at 543 nm using a helium-neon laser.
Fluorescence was detected with a photomultiplier. The samples were placed on a
microscope slide under a glass coverslip.
Rheological measurements of stirred acid milk gels were carried out in a stress-
controlled rheometer (Carri-Med CSL2 100, TA Instruments, UK), using a stainless steel
plate and cone geometry (60 mm diameter, 4 angle). The frequency sweep measurements
were performed from 0.01 to 10 Hz within the linear domain. Tan G corresponds to the
ratio of viscous to elastic properties. All measurements were carried out 1 day after stirred
acid milk gels manufacture.
View Online
3.1 Acid Skim Milk / Guar Gum Gels: Microstructure and Rheological Properties
Figure 1 shows the microstructure and the rheological properties of stirred acid milk gels
with 1.5 wt% of GDL and increasing concentrations of guar gum (0-0.5 wt%). The white
areas indicate the protein stained by DyLight 549, while the dark areas correspond to zones
devoid of protein, thus containing whey and exuded guar gum.
Acid milk gels with no added guar gum were homogeneous. When 0.05 wt% guar
gum was added, the casein micelles tended to form a denser network with larger dark voids
in between (Figure 1a). Those gels exhibited a solid-like behaviour with the G values
higher than G ones, and both G and G showing little dependence on frequency
(Figure 1b). Although filamentous morphology appeared for 0.15 wt% of guar gum, the
G, G and tan G spectra shown in figure 1b and 1c respectively were still that of a typical
structured system but with lower G and G values and slightly higher tan G values than
those obtained for a non-enriched guar gum gel. Increase of guar gum concentration leads
to dramatic changes in stirred gels microstructure and rheological properties. A mix of
filamentous structures and deformed droplets (from 10 to more than 100 m length)
dispersed in a continuous guar gum-rich phase was found for 0.3 wt% guar gum. By
adding a large amount of guar gum to the system (0.5 wt%), protein-rich spherical droplets
were observed. This change in microstructure is accompanied by a clear modification of
the rheological properties of the mixed gels. The solid-like behaviour of the systems tended
to disappear with a loss of connectivity in the network. Instead, a strong frequency
12/04/2014 12:05:23.
dependence of the dynamic moduli (G and G) was observed for those samples indicating
a fluid-like behaviour. Therefore, when increasing guar gum concentrations were added to
Figure 1: (a) Microstructure; (b) Mechanical spectra (G (line) and G (dotted line));
(c) Variation of tan G values with frequency of stirred acid milk / guar gum gels at pH 4.6
containing various guar gum concentrations (0-0.5 wt%) indicated in the figure acidified
at a fixed amount of GDL (1.5 wt%). Microstructure was determined by CLSM. Scale bars
indicate 100 m
View Online
the skim milk system, the mechanical spectra changed significantly with a transition from a
structured system to an entangled polymer solution.
Adding increasing amounts of guar gum to acid skim milk gels induced the creation of
different microstructures. At low guar gum concentrations the stirred gel microstructures
are characterized by a micro-phase separation in the protein network explaining the high
porosity of the network. During milk acidification, the balance between aggregation of
concentrated casein micelles and volume exclusion effects results in compaction of the
casein network and thus to stronger stirred gels.2 Exceeding a given polymer concentration
leads to a phase separation into a protein-enriched phase and a polysaccharide-enriched
phase coexisting in the form of a water-in-water emulsion.7,14,15 At higher guar gum
concentrations (i.e. 0.3 wt%), guar gum forms a viscous continuous phase (as shown in
figure 1b and 1c) containing trapped compact micellar droplets.3
However, for intermediate guar gum concentrations (between 0.15 and 0.3 wt%), the
viscosity of the continuous phase is not sufficient to inhibit droplet mobility. According to
Stokes law, a lower continuous phase viscosity will accelerate the phase separation
process. Therefore the protein-enriched droplets will tend to sediment in mixed gels
containing between 0.15 and 0.3 wt% guar gum, which can explain the elongated and
deformed droplets that were obtained. We will discuss this point more thoroughly below.
Addition of guar gum to skim milk results in a phase separation beyond a given
12/04/2014 12:05:23.
Droplet Sediment
(a) (b) sedimentation layer
25 25
Time (h)
20 0
20
' Backscattering (%)
15 0.1
' Backscattering (%)
0.5
10 0.7 15
1
5
0 10
-5
5
-10
-15 0
0 5 10 15 20 25 30 35 40 0.0 0.2 0.4 0.6 0.8 1.0
Sample height (mm) Time (h)
Figure 2: (a) Delta backscattering values of non-acidified 0.3% guar gum / skim milk
mixture along its height in a glass tube over the time at 43C; (b) Maximum value of delta
backscattering at the bottom of the tube as a function of time. Dotted lines correspond to
the time of the beginning of droplet sedimentation (0.1 h) and apparition of continuous
sediment layer (0.5 h).
View Online
reached after 0.7, 0.5 and 0.2 h with addition of GDL amount of 1, 1.5 and 2.5 wt%
respectively.
Droplet Sediment
sedimentation layer
6.4
6.0
5.6
pH
pH of gel
5.2
formation
4.8
1%
2.5% 1.5%
4.4
0.01 0.1 1
Time (h)
Figure 3: pH kinetics of 0.3 wt% guar gum / milk mixture after various amounts of GDL
addition (1, 1.5, 2.5 wt%). Dotted lines correspond to the time of the beginning of droplet
sedimentation (0.1 h) and apparition of continuous sediment layer (0.5 h) and the pH of
gel formation (pH 5.2) as indicated in the figure. Values are means from triplicate
experiments.
View Online
It seems important to understand the interplay between phase separation and gelation
kinetics (Figures 2 and 3). For instance, in the case of 0.3% guar gum / skim milk mixture
acidified with 2.5 wt% GDL, only a few droplets began to sediment before the pH of the
gel was reached. It should also be noted that with addition of 1.5 wt% GDL the gelation
occurred at the same time as the apparition of the sediment layer. On the contrary there is a
time delay for the system containing 1 wt% GDL to begin phase separation, suggesting
that the phase separation kinetics was quicker than the network formation kinetics.
Relative kinetics of gelation and phase separation seem to govern the morphology of
the mixed systems. The CLSM micrographs in figure 4 show the microstructure of acid
milk / guar gum stirred gels containing 0.3 wt% guar gum and acidified with three different
amounts of GDL (1, 1.5 or 2.5 wt%). Gels prepared with 1.5 wt% GDL exhibited the same
microstructural features as those visualized in figure 1a, composed of a mix of filamentous
structures and deformed droplets dispersed in a continuous guar gum-enriched phase. On
the contrary, small spherically-shaped protein-enriched droplets were found in the system
containing 2.5 wt% GDL, while filamentous structures appeared in the mixed gel acidified
with 1 wt% GDL, i.e with the lowest gelation rate. It should be noted that filamentous
structures and droplets existed before stirring the acid gels and were not oriented after
stirring. As far as we know, such microstructures have never been reported previously for
acid skim milk / guar gum gels.
Upon mixing, the guar gum / skim milk mixture formed an unstable water-in-water
emulsion.7,15 Induced by Brownian motion giving rise to droplet collision, those spherical
domains can either continue to grow through coalescence or/and sediment following
12/04/2014 12:05:23.
Stokes law.7 However, in the case of mixed gels acidified with 2.5 wt% GDL, the gelation
rate was faster than the formation of the sediment layer (Figure 3). Therefore, the small
droplets shown in figure 4 were frozen before complete coalescence mechanism and
sedimentation occurred. Then the gelation process has the effect of freezing the evolution
of the droplets at an intermediate state, far from the thermodynamic equilibrium
conditions, thus hindering additional phase separation process.
For lower gelation rates (addition of 1 or 1.5 wt% GDL to the system), the gelation
process competes with droplet sedimentation and the apparition of a sedimented layer. For
instance, the droplets composing the gel prepared with 1.5 wt% GDL probably had already
grown and coalesced when gelation occurred. However, the sedimentation process is
greatly influenced by the relative densities of the phases, as well as droplet size. Moreover,
owing to very low interfacial tension typical of water-in-water emulsions, droplets are
Figure 4: Microstructure of stirred acid milk / guar gum gels at pH 4.6 containing 0.3 wt%
guar gum acidified with three different GDL concentrations as indicated in the figure (1,
1.5, 2.5 wt%). Scale bars indicate 100 m.
View Online
As previously demonstrated from the microscopic results, it appeared that the dynamics of
protein-enriched droplets in a form of water-in-water emulsion followed Stokes law.
Hence, the parameters that contribute to droplet sedimentation are not only the density
difference between the two phases, the mean droplet size or the viscosity of the continuous
phase, but also gravitation. In order to explore more precisely possible reasons for droplets
extension, guar gum / acid milk gels were prepared in a temperature-controlled centrifuge
using additional acceleration field (2.5 g).
Evidence of droplet spinning due to a higher acceleration field is shown in figure 5b in
comparison to figure 5a. Very long protein-enriched filaments (between 100 and 400 m)
dispersed in a guar gum continuous phase were produced upon centrifugation at 2.5 g
(Figure 5b) whereas smaller filamentous and deformed droplets (between 20 and 100 m)
12/04/2014 12:05:23.
were found in the absence of an additional acceleration field (Figure 5a). Therefore,
increasing the acceleration field produced enough energy to distort the droplets, which had
the effect of making the droplets spin by increasing particle deformation, without
promoting break-up or coalescence. Such structures have been produced in the past using
extensional or orifice flows17, or using appropriate shear flows.14,15 With increasing shear
stress, particle deformation increases which offers a broad flexibility to generate different
extents of anisotropy from deformed droplets to the formation of a string phase.11
As soon as the applied deformation is stopped, the droplet can easily relax to the
equilibrium condition in a spherical form.9 However, if droplet gelation occurs in a non-
equilibrium state, the unstable morphology is trapped by gelation. This phenomenon was
well illustrated using mixtures of gelatin and maltodextrin or dextran14,18,19 for which the
system gelation was induced by decreasing the temperature of the gelatin continuous
Figure 5: Microstructure of stirred acid milk / guar gum gels at pH 4.6 containing 0.3 wt%
guar gum acidified at a fixed amount of GDL (1.5 wt%) while an acceleration field of (a) 1
g or (b) 2.5 g was applied. Scale bars indicate 100 m.
View Online
phase. In the case of acid skim milk / guar gum gels, the triggered freezing of the phase-
separating structure was governed by acid-induced protein aggregation. Thus gelation can
be used to kinetically trap various morphologies created by controlling the particle
dynamics, using modification of the gelation rate (gel formation kinetics) or the
acceleration field (droplet deformation extent).
5 CONCLUSION
The present study has shown how the combined parameters of particle dynamics and
gelation kinetics in acid skim milk / guar gum mixtures can be used to design a wide range
of microstructures. Adding specific amounts of guar gum into skim milk gave rise to a
droplet-like morphology which can be trapped by acid-induced gelation.
A more in-depth experimental investigation of the combined gelation time-phase
separation process was carried out by changing the gelation rate and using appropriate
acceleration field. The potential to generate different types of morphologies, by controlling
the gelation rate and the acceleration field to obtain different droplet deformation extents,
gives a way to manipulate textures. Therefore, future works should focus on quantifying
the droplet dynamics more precisely, which could be achieved through rheo-optical
processes for instance.
Acknowledgements
The results were obtained in the context of Satiarome project which is supported by
12/04/2014 12:05:23.
Vitagora pole, DGCIS and local authorities with the financial support of OSEO and
FEDER. The authors would like to thank Gabrielle Moulin for her assistance with the
confocal microscopy experiments.
References
1 J.L. Doublier, C. Garnier and C. Sanchez, Curr. Opin. Colloid Interface Sci., 2000, 5,
202-214.
2 C. Sanchez, R. Zuniga-Lopez, C. Schmitt, S. Despond and S. Hardy, Int. Dairy J.,
2000, 10, 199-212.
3 D.W. Everett and R.E. McLeod, Int. Dairy J., 2005, 15, 1175-1183.
4 S. Bourriot, C. Garnier and J.L. Doublier, Int. Dairy J., 1999, 9, 353-357.
5 R. Tuinier, E. ten Grotenhuis and C.G. de Kruif, Food Hydrocoll., 2000, 14, 1-7.
6 T.J. Foster, C.R.T. Unerdown, D.P. Ferdinando and I.T. Norton, In: Food Colloids, E.
Dickinson and B. Bergenstahl (Eds.), 1997, 346-356.
7 C. Schorsch, M.G. Jones and I.T. Norton, Food Hydrocoll., 1999, 13, 89-99.
8 G.G Stokes, Trans. Cambridge Phil. Soc., 1851, 9.
9 W.J. Frith, Adv. Colloid Interface Sci., 2010, 161, 48-60.
10 B. Wolf, R. Scirocco, W.J. Frith and I.T. Norton, Food Hydrocoll., 2000, 14, 217-225.
11 B. Wolf, W.J. Frith and I.T. Norton, In: Gums and Stabilizers for the Food Industry
11. P.A. Williams, G.O. Philipps (Eds.), IRL Press, Cambridge, 2002, 112-119.
12 P.A. Aichinger, M.L. Dillman, S. Rami-Shojaei, M. Michel and D.S. Horne, In: Food
Colloids, E. Dickinson and M.E. Leser (Eds.), 2006, 283-296.
13 D.S. Horne, In: Food Colloids, E. Dickinson and R. Miller (Eds.), 2001, 345-351.
14 I.T. Norton and W.J. Frith, Food Hydrocoll., 2001, 15, 543-553.
15 V.B. Tolstoguzov, Food Hydrocoll., 2003, 17, 1-23.
16 J.A. Lucey, J. Dairy Sci., 2002, 5, 281-294.
View Online
17 Y.A. Antonov, V.Y. Grinberg, N.A. Zhuravskaya and V.B. Tolstoguzov, J. Texture
Stud., 1980, 11, 199-215.
18 N. Lorn, M. Langton and A.M. Hermansson, Food Hydrocoll., 1999, 13, 185-198.
19 P. Van Puyvelde, Y.A. Antonov and P. Moldenaers, Food Hydrocoll., 2003, 17, 327-
332.
12/04/2014 12:05:23.
COMPRESSION TEST OF FOOD GELS ON AN ARTIFICIAL TONGUE AND ITS
COMPARISON WITH SENSORY TESTS
Sayaka Ishihara1, Mai Isono1, Satomi Nakao1, Makoto Nakauma1, Takahiro Funami1,
Kazuhiro Hori2, Takahiro Ono3, Kaoru Kohyama4 and Katsuyoshi Nishinari5
1
San-Ei Gen F.F.I., Inc.,1-1-11 Sanwa-cho, Toyonaka, Osaka 561-8588, Japan
2
Niigata University Graduate School of Medical and Dental Sciences, 2-5274 Gakkocho-
dori, Chuo-ku, Niigata 951-8514,Japan
3
Osaka University Graduate School of Dentistry, 1-8 Yamada-Oka, Suita, Osaka 565-0871,
Japan
4
National Food Research Institute, National Agriculture and Food Research Organization,
2-1-12 Kannondai, Tsukuba, Ibaraki 305-8642, Japan
5
Glyn O. Phillips Hydrocolloid Research Center, School of Food and Pharmaceutical
Engineering, Faculty of Light Industry, Hubei University of Technology, Wuhan 430068,
China
12/04/2014 12:05:25.
1 INTRODUCTION
The human tongue plays a crucial role throughout food oral processing, including
recognition, transportation to the molar or cheek teeth, compression or squeezing against
the hard palate (i.e. tongue-palate compression), mixing of food particles with saliva (i.e.
bolus formation), and transportation of the bolus to the pharynx. Among a series of oral
strategies, size reduction should be focused more extensively for the texture design of food
products because there is an increasing number of elderly people and patients with
masticatory disturbance in recent times.
In our previous study, an in vitro evaluation system of food texture was developed
using a combination of artificial tongue and a conventional uni-axial compression
apparatus to mimic tongue-palate compression1. Artificial tongues were prepared from
silicone rubber with different rheological properties by changing silicone rubber
concentration to make apparent elastic moduli of artificial tongues equivalent to those of
human tongues from a relaxed to a tension state. Agar gels of comparable fracture strain
but different fracture forces were used as a model food. It was concluded that the fracture
profile of agar gel on artificial tongue during compression with a non-deformable
aluminium platen approximated to the choice of the oral strategy for size reduction when
apparent elastic modulus of artificial tongue was approx. 5.5104 Pa. That is, agar gels
which fractured upon compression on the artificial tongue were consumed by tongue-
View Online
palate compression, whereas the gels which did not fracture on the artificial tongue were
consumed by teeth mastication.
In the present study, gels from gellan gum were prepared over a wide mechanical
range and evaluated on the same evaluation system to validate our previous conclusion; to
confirm the robustness of the evaluation system and also the limitation if there is, followed
by the deduction of possible causes for exception.
As an artificial tongue, silicone rubbers of different apparent elastic moduli were prepared
using a room temperature vulcanization (RTV) rubber (KE-12, Shin-Etsu Chemical Co.,
Ltd., Tokyo, Japan). The KE-12 was mixed with silicone oil (RTV thinner, Shin-Etsu
Chemical Co., Ltd.) at ratios of 4:6, 5:5, and 6:4, onto which a curing catalyst (CAT-RM,
Shin-Etsu Chemical Co., Ltd.) was added at 0.5% (w/w) at room temperature. After
deaeration under vacuum, the mixture was placed into cylindrical glass molds of 20 mm in
diameter and 10 mm in height and cured at 20C for 2 days. The silicone rubbers obtained
were termed as from S40 to S60 in the increasing order of silicone rubber concentration.
Apparent elastic moduli of these artificial tongues determined by the stress/strain ratio at
20% strain through compression were 1.83104, 5.49104, and 11.30104 Pa,
corresponding to that of the human tongue in a relaxed state (1.22104 Pa) and in a tension
state (12.25104 Pa).
12/04/2014 12:05:25.
As a model food, gellan gels of various fracture strains and forces were prepared using a
mixture of KELCOGEL (low-acylated gellan gum), and KELCOGEL LT-100 (high-
acylated gellan gum) (both from San-Ei Gen F.F.I., Inc., Osaka, Japan). To mask subtle
flavour from these polysaccharides, which may affect the results from sensory evaluation,
sucrose was added at 10% (w/w) to all gel samples. A mixture of sucrose and gellan gum
was added to de-ionized water at 90 C in glass beakers of 500 mL and stirred at 1,300 rpm
for 10 min at the same temperature, then calcium lactate (0.1% as pentahydrate) and a
food colour (0.2%, SAN GREEN GC-EM, San-Ei Gen F.F.I., Inc., Osaka, Japan) were
added to the solutions. Solutions obtained were poured into plastic cups of 65 mm in
diameter and 25 mm in height, inside which cylindrical glass molds of 20 mm in diameter
and 10 mm in height were already placed. The solutions in the cups were hermetically
sealed, heated at 85C for 30 min, and refrigerated at 8C for 2 hours. The gels obtained
were subjected to sensory evaluation and mechanical test at 20C after curing at 5C
overnight. Fracture strain and force of gellan gels are shown in Figure 1. It should be
noted that these determinations were from compression with a metal probe on a hard stage
(not on artificial tongue). The gels were termed after their fracture strains (from series A
to series D in the increasing order) and fracture forces (from approx. 10 to 30 N in 5 N
increments).
View Online
A B C D
35 Low- High-
30
30 acylated acylated
gellan gellan
25
Fracture force (N)
25
A 100 0
20 20
B 80 20
15
15 C 70 30
10 10 D 50 50
5
Sample denomination
0
40 50 60 70 80
A 10
Fracture strain (%)
Mixing ratio of two Fracture
gellan gums force
2.3 Sensory evaluation asking oral strategy for the first size reduction
Five males and five females ranging from 25 to 41 in age (31.7 years old on average) and
having normal dental status participated in this experiment. They were all trained for
12/04/2014 12:05:25.
evaluating gel texture. Gellan gels of 20 mm in diameter and 10 mm in height were served
at 20C. Subjects were allowed to process the gels orally without any restriction. After the
process, subjects were asked which oral strategy was used for the first size reduction;
tongue-palate compression or teeth mastication. Evaluation was done in 4 separate
sessions starting from series A to series D and repeated twice for each subject. Choosing
the probability of tongue-palate compression was determined using 20 data.
All subjects used tongue-palate compression for gels of relatively smaller fracture forces
(10 and 15 N) regardless of the fracture strain (Figure 2). The number of subjects who
used teeth mastication increased with increasing fracture force, which was more apparent
for gels of relatively larger fracture strains (series D rather than series A), and equal to or
more than 50% of the subjects used teeth mastication for A30, B25-30, C25-30, and D25-
30. This suggests that a force threshold during the first size reduction to change oral
strategy from tongue-palate compression to teeth mastication should decrease with
increasing fracture strain of the gels.
Deformation and fracture profiles of gellan gels were observed during mechanical
compression on each artificial tongue. The size was the same between artificial tongue and
gellan gel (i.e. 20 mm in diameter and 10 mm in height). The combination (upper: gellan
gel, bottom: artificial tongue) was compressed uni-axially with an aluminium platen of 50
mm in diameter at a crosshead speed of 10 mm/s or 5 mm/s up to 50% strain of the
combination at 20C (Figure 3(a)). Deformation behaviour till fracture was pictured by a
digital camera Everio (GZ-V570, JVC KENWOOD Corporation, Kanagawa, Japan) as a
movie, which was synchronized with the stress-strain curve, ensuring the fracture point.
Snap shot images captured from the movie are presented in Figures 3(b) and (c). Fracture
View Online
Figure 2: Percentage of oral strategy choice and fracture probability of gellan gels on
each artificial tongue a crosshead speed of 10 mm/s
(a) Series A (fracture strain approx. 45%), (b) series B (fracture strain approx. 55%-
60%), (c) series C (fracture strain approx. 60%-65%), and (d) series D (fracture strain
approx. 70%-75%).
View Online
For series A to series C, results from sensory evaluation related the best to the fracture
profile of gellan gels on S50 artificial tongue, and choosing the probability of tongue-
palate compression as an oral strategy for the first size reduction corresponded to the
fracture probability of the gels on S50. For gellan gels which did not fracture on S50,
12/04/2014 12:05:25.
equal to or more than 50% of subjects used teeth mastication instead of tongue-plate
compression for the first size reduction. For series D, on the other hand, results from
sensory evaluation related to the fracture profile of gellan gum on S60 better than that on
S50. Also, when crosshead speed was decreased from 10 to 5 mm/s in the mechanical test,
results from sensory evaluation related well to the fracture profile of gellan gels even on
S50 (Figure 4).
Figure 4 Percentage of oral strategy choice and fracture probability of gellan gels (series
D) on artificial tongue S50 at a crosshead speed 5 mm/s
View Online
3 CONCLUSION
The oral strategy for the first size reduction has been simulated mechanically using a
combination of an artificial tongue (apparent elastic modulus of approx. 5.5104 Pa) and a
conventional uni-axial compression apparatus. Conclusion from a previous study on the
same evaluation system is applicable to series A to series C (approx. 45%-65% fracture
strain) but not to series D (approx. 70%-75% fracture strain). In the case of series D, it is
suggested that tongue-palate compression should be performed during the first size
reduction at higher tongue excitation and/or at lower compression speed. In addition,
human subjects may apply shear to these gels for size reduction, which should be
incorporated in future work.
Acknowledgment
This study was partially supported by the research and development projects for
application in promoting new policy of Agriculture Forestry and Fisheries 22026.
References
1. INTRODUCTION
There is a growing interest in the food industry for the use of submicron emulsions. They
present several advantages over conventional emulsions that include an increase in the
bioavailability of lipophilic components, greater stability to aggregation and gravitational
12/04/2014 12:05:27.
separation, higher surface area for controlled release, and as a result an improved
commercial shelf life1, 2.
Submicron emulsions are usually formed with low molecular weight surfactants and little
work has been conducted with the use of proteins as emulsifiers for these systems2. The
use of proteins for the formation of submicron emulsions offers many benefits such as
cleaner labelling than surfactants and added nutritional value3.
Proteins are capable of stabilising oil-water interfaces due to their surface active nature.
Their surface activity is due to their amino acid sequence which produces hydrophobic and
hydrophilic regions. At the oil-water interface proteins adapt to the most entropically stable
state, where the hydrophilic and hydrophobic amino acid regions protrude into the aqueous
and oil phases respectively with the state of least energy. Proteins form a thick interfacial
layer due to hydrophobic interactions between adsorbed protein and protein in the bulk4
which prevents droplet coalescence. Proteins are also charged molecules which, when
adsorbed at the interface reduce the likelihood of droplet collision due to electrostatic
repulsion, producing stable emulsions3, 5, 6.
NaCas (sodium caseinate), WPI (whey protein isolate) and MPI (milk protein isolate)
present different structures and sizes which influence their performance as emulsifiers.
WPI is a larger protein than NaCas3 and as a result it takes longer to diffuse through the
bulk phase to adsorb at the interface4. Coupled with this a globular protein such as WPI
undertakes a long conformational denaturation process to stabilise oil-water interfaces4. As
a consequence NaCas is expected to adsorb at the interface more rapidly than WPI during
emulsification, which may affect emulsion physical properties3.
View Online
Submicron emulsions can be produced using various methods such as high energy devices
(high pressure homogeniser or microfluidiser)7 or chemical methods (phase inversion
temperature or composition)8. This research has been conducted primarily with the use of
surfactants2, 9. Research in the area of submicron emulsions stabilised with proteins with
proteins is limited. The incorporation of an ingredient in the oil phase which is volatile can
be evaporated by rotary evaporation so as to reduce the droplet size. This technique has
been used with whey protein to further reduce the droplet size after homogenisation9, 10.
In this work, oil-in-water submicron emulsion formation was conducted with different
concentrations of native and sonolysed dairy proteins and a low molecular weight
surfactant for comparative purposes. A direct method of emulsification (high pressure
homogenisation) was used for emulsion formation. The molecular weight profile of the
dairy proteins was reduced using sonolysis. Droplet formation and emulsion stability of
emulsions produced with these hydrolysates were compared to native dairy proteins. By
studying a range of dairy proteins and changing their molecular weight we seek to
demonstrate the important role that protein structure and properties play in the formation
and stabilisation of submicron emulsions.
Materials:
Tween 80, purchased from Sigma Aldrich (UK), as well as acid casein (A290), whey
protein isolate (W994) and milk protein isolate (A055) all provided by Kerry Ingredients
(Listowel, Ireland) were used as emulsifiers as part of this study. The oil used was
commercially available pure vegetable oil. NaCas was prepared from acid casein using a
method described by OConnell and Flynn3. WPI, MPI and Tween 80 were dissolved in
water at their native pH values.
Sonolysis:
Protein hydrolysates were prepared by sonicating protein solutions for two minutes with an
amplitude of 95% using an ultrasonic probe (Viber Cell 750, Sonics, USA).
Emulsification:
Pure vegetable oil was added to a protein/surfactant solution at different concentrations.
This mixture was emulsified at 8000 rpm for 2 minutes in a high shear mixer (SL2T,
Silverson, UK) to form a pre-emulsion. This pre-emulsion was passed through a high
pressure homogeniser (Panda NS 1001L-2K, GEA Niro Soavi, UK) at 1,250 bar for 2
passes. The concentration of emulsifiers that were used as part of this study were 0.1%,
0.5%, 1%, 5% and 10%. The mass fraction of the oil phase was kept constant at 10% for
all samples.
Protein Size:
Protein size was measured by dynamic light scattering (Zetasizer Nano Series, Malvern
Instruments, UK). In order to avoid turbid samples, protein solutions were diluted to a
concentration of 0.1% for the measurement of size.
View Online
Intrinsic Viscosity:
Protein solution viscosity was measured using Kinexus rheometer (Malvern Instruments,
UK) equipped with a double gap geometry at a constant shear rate of 250 s-1. Intrinsic
viscosity was then extrapolated by using the Kraemer and Huggins equations11.
emulsifiers at concentrations below 1%. Proteins are charged molecules and electrostatic
repulsion ensures that droplets do not come into contact with one another 13. Conformation
of protein at the interface allows steric stabilisation of droplets. Neither of these effects
occur for droplets stabilised by Tween 80. This is due to the non-ionic nature and low
molecular weight of Tween 80. This results in the formation of larger emulsion droplets for
Tween 80 than emulsions droplets stabilised by NaCas and WPI.
Table 1: Effect of concentration of Tween 80, NaCas, WPI and MPI on the formation and
stability, at day 28, on the d3,2 (nm) of emulsions at concentrations of 0.1%, 0.5%, 1%, 5%
and 10% % with a comparison between native and sonolysed proteins.
At a concentration of 0.1% the difference between NaCas and WPI stabilised emulsions
droplets is ~200 nm. This is due to the relative size and the nature of the protein (flexible
or globular). These factors influence the rate at which they adsorb to the oil-water
interface. WPI is larger than NaCas and as a result it takes longer to reach the interface due
to reduced molecular mobility. The rate of adsorption is affected by the structure of the
protein. WPI as a globular protein takes longer to adsorb to the interface due to the
conformational denaturation necessary to stabilise the droplets. These are the reasons why
NaCas produces smaller droplets at a concentration of 0.1% than WPI.
At MPI concentrations below 1%, large droplets ~3 m are formed. MPI is composed of
~80% casein micelles and ~20% whey protein. Micellar casein in MPI are similar to solid
particles with a size range of 200 1000 nm14 which are incapable of stabilising submicron
droplets due size and hydrophobicity15. At low concentrations, the large droplet size is
likely due to the low amount of emulsifier, i.e. whey protein, which cannot ensure an
efficient surface coverage of the submicron droplets.
Sonolysis has been shown to reduce the molecular weight profile of proteins through
intrinsic viscosity. Figure 1 shows the Huggins and Kraemer extrapolations for the
determination of the intrinsic viscosity of both native and NaCas which has been sonolysed
and intrinsic viscosity values for native and sonolysed proteins are given in Table 2.
12/04/2014 12:05:27.
1.4
1.2
Intrinsic Viscosity (dl/g)
0.6
0.4
0.0 0.1 0.2 0.3 0.4 0.5
Concentration (w/v%)
Figure 1: Comparison of the intrinsic viscosity values for native and NaCas which has
been sonolysed for 2 minutes using the Huggins-Kraemer extrapolation method.
View Online
Table 2: Comparison of the intrinsic viscosities (dl/g) of native and sonolysed NaCas, WPI
and MPI.
The reduction in the intrinsic viscosity values between native and sonolysed protein shows
that there is a reduction in the molecular weight. This reduction in molecular weight can be
estimated with the Mark-Houwink equation, however the parameters of this equation (K
and ) need to be evaluated.
Figure 2 shows structural differences of 5% NaCas solution before and after sonication.
Native proteins (Figure 2a) form aggregates of uniform sizes that assemble in a dense
network, while sonicated protein aggregates (Figure 2b) are more polydispersed and
discrete entities. This is due to the cleavage of the peptide bonds after sonication which
reduces the molecular weight profile of the protein.
Emulsion droplet size for both sonolysed NaCas and WPI is the same as native NaCas and
WPI at the concentrations tested. Sonicated MPI results in the formation of different
droplet sizes by comparison to native MPI. At concentrations below 1%, a decrease in
droplet size is observed with using sonolysed MPI, from ~3000 nm to ~6000 nm. This is
likely due to casein micelle break-up during sonication resulting in free casein molecules
12/04/2014 12:05:27.
Figure 2: Comparison of the effects of sonication on 5% NaCas. (a) shows NaCas before
sonication and (b) after sonication for 2 minutes. Scale bar is 500 nm.
View Online
4. CONCLUSIONS
Formation of stable submicron emulsions was achieved with the use of native and
sonolysed dairy proteins by direct methods of homogenisation. Droplet sizes are
comparable to that of emulsions formed with a low molecular weight surfactant. Emulsion
droplet size and stability were dependant on both concentration and protein type at
concentrations 1% and below. At concentrations above 1%, the droplet size was the same
and was not dependant on the type of protein. Native and sonolysed dairy proteins
produced emulsions with similar droplet sizes with the exception of MPI where sonolysed
MPI resulted in the formation of smaller droplets
than native MPI due to the disruption of the micellar casein by sonication which allows for
more rapid absorption to the interface than with native MPI. Sonolysis has been shown to
reduce the molecular weight of proteins with the use of intrinsic viscosity.
Acknowledgements
The authors wish to thank Kerry Group for their sponsorship and permission to publish this
work. The authors also thank the EPRSC for financial support.
References
12/04/2014 12:05:27.
F.A. Husband1, M.J. Ridout1, P.S. Clegg2, M. Hermes2, J. Forth2, W.C.K. Poon2 and P.J.
Wilde1
1
Institute of Food Research, Norwich Research Park, Norwich NR4 7UA, UK.
2
School of Physics & Astronomy, University of Edinburgh, Mayfield Road, Edinburgh
EH9 3JZ, UK
1 INTRODUCTION
Obesity and the associated health risks are amongst the main health challenges facing the
UK. Despite increased awareness of a healthy diet and lifestyle, the number patients in the
UK diagnosed with obesity is increasing year on year. Weight gain in most individuals
12/04/2014 12:05:29.
tends to be long term, resulting in an increase in BMI with age due to a small positive
energy balance1 showing a maximum in BMI for the 55-64 age group. This is thought to
be due to a combination of a small excess energy intake and reduced energy expenditure, 1
in combination with our ancient evolutionary drivers for a positive energy balance to store
energy to tide us through times of low food availability.2 Therefore small, but long term
reductions in energy intake could offset this gradual life-long increase in weight and stem
the rising incidence of obesity. However, our inherent drive to overconsume in order to
store energy for lean times, means that individuals have to make a conscious decision to
either reduce energy intake or increase energy expenditure. To satisfy consumers desires
for taste and quality, but in healthier, reduced energy form, a wide range of reduced calorie
foods have been developed over many years. Some have been successful, and, in the case
of semi skimmed milk, is now the norm for many consumers. However, for certain classes
of foods, reducing the energy content of the food without adversely affecting its perceived
quality is more challenging.
Reduced fat versions of many emulsified foods such as dairy products, dressings, sauces,
ice creams etc are often perceived as poorer quality because the perception of fat in food is
a complex mixture of sensory (taste, texture and aroma), physical and psychological
factors.3 Fat content and oil droplet size both have an effect on perceived creaminess of
foods; a higher fat content and smaller oil droplets both give a perception of increased
creaminess,3 a characteristic often attributed to a pleasant eating experience. The
perception can thus be influenced by the fundamental physical or colloidal properties of
the emulsion. Many reduced fat foods are perceived as thinner and less viscous and
require other interventions to replace the fat, resulting in more additives, which are disliked
by consumers due to their non natural appearance on labelling. Many foods contain
emulsified fats to impart desirable textures and flavours, and can account for up to 25% of
our dietary fat intake.
View Online
Oil in water emulsions were prepared using commercial sunflower oil as the dispersed
phase. A phase volume of 0.3 was used throughout the study. Protein stabilised emulsions
were prepared using a whey protein isolate (WPI) (Bipro, Davisco), either untreated, or
heat treated. WPI solutions were prepared at 1 wt% in 10 mM phosphate buffer, pH 7.
Heat treated WPI was prepared by heating the WI solution in a water bath at 70C for 30
minutes. The protein stabilised emulsions were compared to those emulsified by a low
molecular weight surfactant blend. A 1:1 mixture of Brij35 and SDS was prepared and
View Online
used at a concentration of 0.1 wt%. The mixture of ionic and non-ionic surfactant resulted
in a zeta potential which was close to that of the WPI emulsion.10 Emulsions were prepared
at a range of droplet sizes by varying the energy of homogenisation. 60 % sunflower oil in
water emulsions were initially prepared using a Waring laboratory blender at low, medium
or high power for 1 minute. These emulsions were further emulsified using a high pressure
homogeniser (Emulsiflex B3, Avestin, Toronto). The homogenisation pressure was varied
between 1,000 and 2,500 psi to create fine emulsion droplets.
The droplet size distribution and specific surface area of the emulsions were determined
using a laser diffraction particle sizer (LS13-320, Beckman-Coulter). The viscosity of the
emulsions was increased through the addition of a stabilising hydrocolloid solution either
xanthan (Keltrol, CP Kelco) or a cold swelling starch (pre-gelatinized waxy maize starch,
Ultratex, Univar Ltd). In order to ensure that the viscosity of the hydocolloid was not
affected by the homogenisation process, the hydrocolloid was prepared at double the
required concentration, then gently mixed with the emulsion using an overhead stirrer for 1
hour to ensure homogeneous mixing.
2.2 Rheology
The viscoelastic modulus of the emulsions was measured using a AR-G2 controlled stress
rheometer (TA Instruments). All emulsions displayed non-Newtonian shear thinning
behaviour. The viscoelastic modulus was monitored as a function of time at an amplitude
of 0.1 %, which was within the linear response regime of the viscoelastic modulus. Figure
1 shows typical results for visco-elasticity for the different emulsifiers used in 0.3 wt%
12/04/2014 12:05:29.
xanthan as a function of time. At this droplet size, clear differences in the behaviour of the
emulsions were observed. The surfactant stabilised emulsions typically reached an
equilibrium elastic modulus after 60 minutes, after which the modulus remained constant.
In contrast, the protein stabilised emulsions tended to take much longer to equilibrate,
particularly the heat treated WPI samples, which would continue to increase for several
hours. This was probably due to some reorganisation of the emulsion structure, possibly
due to interactions between the adsorbed protein and the xanthan matrix or flocculation.
However, extensive flocculation was not observed in any of the emulsions using
microscopy or light scattering, but this does not discount low level flocculation events
which may lead to a strengthening of the hydrocolloid/emulsion network.
Figure 2 shows the impact of droplet size on the elastic modulus of emulsions stabilised by
the different emulsifiers. The modulus was found to increase in a linear fashion as a
function of specific surface area. So the data is presented as a function of the emulsion
specific surface area (SSA) to account for the difference in absolute droplet size
distributions between the emulsions. The elastic moduli of the emulsions increased with
increasing SSA in a broadly linear fashion for the surfactant and untreated WPI results not
shown.
There was a clear difference in the slope of the relationship, with WPI emulsions
displaying consistently higher values than the surfactant, and the difference being greater
at higher SSA. Therefore to enhance the effect of the emulsifier, the smallest droplet size
distribution is preferred. The key difference in interfacial properties between the protein
and surfactant is interfacial rheology/mobility. WPI forms a cohesive elastic network at an
interface, whereas these surfactants form a fluid, mobile layer on the droplet surface.10 The
heat treated WPI emulsions led to further increases in the emulsion elastic moduli,
particularly after long times due to the observed continued time dependent increase in the
modulus observed in Figure 1. This led to a non-linear relationship between the elastic
modulus and SSA, with much enhanced values of the elastic modulus at the higher SSA
values. The mechanisms underlying these results are not clear, but probably goes beyond
an effect of the interfacial rheological properties. The heat treatment leads to some
aggregation of the WPI due to partial unfolding and exposure of hydrophobic groups. Too
12/04/2014 12:05:29.
much aggregation leads to diminished functionality as the aggregates become too large to
diffuse / adsorb to the interface quickly enough to stabilise the emulsion during
homogenisation.13 This led to a decrease in the SSA for samples treated at higher
temperatures or for longer times (results not shown). The heat treatment led to limited
aggregation, and was observed to lead to a rougher interface on the surface of the
droplets, which was revealed by TEM (results not shown). This increased adsorbed layer
thickness and roughness could lead to enhanced interactions between the adsorbed layer
and the xanthan matrix, or effectively increase the dispersed phase volume.
View Online
2.3 Rheo-Imaging
Rheo-Imaging was studied using confocal microscopy mounted in situ beneath a modified
rheometer with a glass base plate to allow direct observation of the emulsions under shear.
Images were acquired at a rate of up to 100 s-1 during oscillatory shear. The emulsions
were matched for refractive index to allow clear imaging of droplets deep into the
emulsions. Figure 3 shows snapshots of images from either end of an oscillation showing
how the spatial position of individual droplets can be tracked as a function of amplitude.
The displacement velocity of the droplets can then be plotted as shown. A clear difference
between the behaviour of protein and surfactant stabilised emulsions is observed, showing
a difference in the velocity of the droplets close to the base plate at higher strains,
suggesting a greater degree of slip in the surfactant emulsion compared to the protein
emulsion.
Two types of thickener were used in the preparation of the emulsions, a cold swelling
starch and xanthan. This was to test how the morphology of the emulsions and gel matrix
can influence the rheological behaviour. Xanthan forms an entangled polymer network,
whose molecular dimensions are in the order of 10 nm, smaller than that of the emulsion
droplets. In contrast, the starch forms a network of swollen starch granules around 50 Pm
in diameter, much larger than the emulsion droplets as shown in Figure 4.
View Online
Figure 4: Left, effect of adding WPI or emulsions to xanthan or starch gels. Right, optical
micrograph of starch-emulsion mixture, showing large starch granules in blue, separated
by emulsion droplets.
The effect of the interfacial composition on the rheological properties of the emulsion are
shown in Figure 4. Interestingly although the effects of adding the emulsions are different
between the two thickeners, the relative effects of the protein vs surfactant emulsions are
12/04/2014 12:05:29.
very similar. Addition of the emulsions to the xanthan gel caused an overall increase in the
modulus, whereas for the starch, the emulsion modulus is reduced when an emulsion is
present. However, in both cases, the modulus of the protein stabilised emulsions is
significantly higher than the surfactant stabilised ones.
3 DISCUSSION
Figure 2 shows the effect of the droplet size distribution and for all systems the emulsions
G' increases as the droplet size decreased, as expected, since the number of droplets
increases and the droplet interactions, which contribute to the rheological properties of the
emulsion will be proportional to the total surface area available for interaction. The reason
for looking at a range of droplet sizes was to account for differences in absolute droplet
size distribution between samples of different interfacial composition. In a previous work
we displaced a protein stabilised emulsion with a surfactant so that we could directly
compare two emulsions with identical droplet size.10 The slope of the G' vs SSA showed
that the protein emulsions yield a higher increment in G' as SSA is increased.
The main difference in interfacial properties between the protein and surfactant is that the
WPI forms an immobile, elastic interface and the surfactant a fluid surface.10 The protein
droplets may be less deformable, but this is likely to be size dependent, as very small
droplets tend to behave like rigid particles as the size decreases. The hydrodynamic
interactions between the water phase and the oil phase will be strengthened in the protein
View Online
case, such that the strain applied to the water phase will be transmitted to the oil phase.
Whereas for a fluid interfacial layer, the motion of the aqueous phase is absorbed by the
motion of the fluid layer and the energy is not transmitted to the oil phase. Similarly, a
rigid macro molecular adsorbed layer is more likely to be able to interact, if only on a
hydrodynamic level, with the polymer matrix. This could result in the emulsion droplets
making a greater contribution to the emulsion G', behaving as 'active' filler particles, as
described previously.14 Surfactant coated droplets behave as 'inactive' filler particles, and
contribute less to the total G' of the system. Additional measurements using a surfactant
with a melting point transition at 42C showed that at temperatures <42C, where the
surfactant behaves like a solid layer,15 the emulsion G' is very close to the unheated WPI,
when the temperature was increased to >42C, when the surface is fluid,15 the emulsion G'
decreases towards that of the surfactant. Thus the unheated WPI emulsion rheology seems
to depend mainly on the interfacial rheological behaviour,
In the case of the heated WPI, the response becomes non-linear as the droplet size
decreases, suggesting some additional forces were contributing to the overall G' of the
system. Heating stimulates aggregation of the WPI due to the partial unfolding of the
protein and subsequent exposure of hydrophobic groups. As discussed earlier, as long as
the aggregates are not too large to affect their adsorption to the interface, stable emulsions
can still be created. The aggregation resulted in a thicker, and rougher layer as evidenced
by separate TEM images. This could lead to greater hydrodynamic interactions with the
12/04/2014 12:05:29.
aqueous phase matrix, due to the increase in surface area resulting from the greater
roughness. In addition, an increased adsorbed layer thickness will also effectively increase
the dispersed phase volume. Simple calculations showed that for the size ranges relevant
to this study, small increases in adsorbed layer thickness, in the order of a few nm could
lead to a significant increase in the apparent dispersed phase volume. Therefore the
combination of increased layer volume and increased surface roughness of the heated WPI
adsorbed layer may well explain the further, non-linear increases in emulsion G' compared
to the unheated WPI.
Changing the scale of interactions between the oil droplets and the hydrocolloid matrix
was investigated. As described above, the xanthan matrix is a molecular network with
typical length scales in the 10 nm range. Figure 4 shows that the presence of an emulsion
strengthens the system. In contrast, for the starch, which is formed from large starch
granules is weakened by the presence of the emulsion, in either the protein or surfactant
case. For the xanthan, the oils droplets appear to act as filler particles, adding further points
of interaction across the network, and maybe even forcing the xanthan molecules into
closer contact by convoluting their path around and in between the droplets. The
interaction between the droplets and the xanthan, as discussed earlier appears to be
stronger for the protein covered droplets compared to the surfactant covered droplets.
However, for the starch, the network relies on interactions over scales much larger than the
emulsion droplets, so the presence of the droplets interferes with those interactions, as the
interaction between the droplets and the starch is much weaker than the interaction
between neighbouring starch granules. Nevertheless, the presence of the protein emulsion
still results in a stronger network than the surfactant stabilised droplets. Hence the
interfacial properties of the droplets is still influencing the strength of the whole network.
View Online
4 CONCLUSIONS
5 REFERENCES
1
Department of Chemical and Biological Sciences, University of Huddersfield,
Queensgate, Huddersfield HD1 3DH, UK
2
Department of Food Technology, ATEI of Thessaloniki, 57400 Thessaloniki, Greece
1 INTRODUCTION
Pectins are widely utilized in the food industry for their gelling, stabilizing and thickening
properties.
High emulsification capacity is usually attributed to proteins whereas
polysaccharides typically demonstrate negligible surface activity at the o/w interface due to
their hydrophilic character and are, therefore, not so useful as emulsifying agents.
Similarly to most polysaccharides, pectins are not normally considered as emulsifying
agents except for cases such as the acetylated pectin from sugar beet that has been shown
to possess greater surface activity than commercially produced low- or high-methoxyl
pectins and are capable of producing and stabilizing fine o/w emulsions1. Citrus pectins
with low molecular weight of about 60-70 g mol-1 and high degree of methoxylation were
found to be good emulsifying agents 2 whereas the emulsifying properties of sugar beet
pectins were attributed to the presence of acetyl groups (45%), co-extracted protein
fraction and ferulic acid moieties covalently attached to pectin molecule 1, 3, 4.
Okra pectins are found to be acidic, random coil polysaccharides composed of
galactose, rhamnose and galacturonic acid. The repeating unit was reported to be -(1-2)-
rhamnose and -(1-4)-galacturonic acid residues including disaccharide side chains 5 and
they form viscous solutions that exhibit pseudo-plastic behavior 6-8. They differ greatly
from those extracted from apple, citrus and beet in terms of protein and acetyl contents,
indicating their greater hydrophobicity and therefore substantial surface activity at the o/w
interface 9-12 suggesting that pectin extracts derived from okra can be used as an effective
emulsifying agent.
The aims of present work, therefore, were to obtain okra extracts rich in pectins
using different extraction protocols and determine their emulsifying capacity in model o/w
emulsions under acidic conditions.
View Online
Extraction of okra AIS (Alcohol Insoluble Solids) has been performed as described
previously7, 13. Samples extracted at pH 6.0 or 4.0 were labeled OE6 or OE4,
respectively.
The samples of a concentration 2.5 mg mL-1 were eluted with ultrapure water
containing 0.02% NaN3 with a flow rate of 1 mL min-1. The HPSEC system comprised of
(i) a SpectraSystem SCM 1000 degasser (Thermo Separation Products, San Jose, CA); (ii)
a SpectraSystem P 2000 chromatographic pump (Thermo Separation Products, San Jose,
CA), (iii) a 2 m frit (Idex, Oak Harbor, WA); (iv) a GPC/SEC PL-Aquagel-OH 507.5
mm guard column (8 m) (Varian Inc, Palo Alto, CA); (v) two tandem GPC/SEC PL-
Aquagel-OH 3007.5 mm columns (Varian Inc, Palo Alto, CA); (vi) an ERC 7515
refractive index detector (Rigas Labs, Thessaloniki, Greece). Data acquisition and handling
were performed with EZChromTM software (Scientific Software Inc, Pleasanton, CA).
The aqueous phases of the emulsions were prepared by means of dissolving okra extract
12/04/2014 12:05:31.
powders at 1.8% w/v concentration in citric buffer (10 mM, pH 3.0) at room temperature.
Okra extract solutions were characterized at 1.8% w/v concentration. For emulsion
preparation, the above aqueous phases were magnetically stirred with hexadecane for 3
min in order to produce emulsion pre-mixes with M = 0.2 and 1.5% w/v final extract
concentration in the entire emulsion. This pre-mix was immediately homogenized (IKA
T18 basic, Ultra-Turrax, Germany) for 1 min. For the determination of long-term stability
all emulsions were stored in an incubation chamber at 25qC.
Droplet size distribution was measured immediately after the emulsion preparation and
after 5, 10, 20 and 30 days of storage using a Malvern Mastersizer 2000 (Malvern
Instruments Ltd, Worcestershire, UK) laser diffraction particle size analyzer using the
small volume sample dispersion unit Hydro 2000SM (Malvern Ltd, UK) while dispersion
was performed on an appropriate buffer, similar to those of the emulsions. Refractive index
of hexadecane and dispersion medium (citric buffer, 10 mM, pH 3.0) was set to 1.435 and
1.333, respectively. The measurements were performed in duplicates in three different
emulsion preparations yielding in total six replicates for each sample.
Rheological properties of samples were measured using a Bohlin Gemini 200HR Nano
rotational rheometer (Malvern Instruments, Malvern, UK) equipped with cone-and-plate
geometry (40 mm diameter, cone angle 4q) and Peltier temperature controller. All
measurements were performed in a steady shear mode in the range 0.011000 s-1 at 25 qC.
Viscosity measurements were conducted immediately after preparation of okra extract
View Online
solutions and emulsions and after 5, 10, 20 and 30 days of storage. All measurements were
performed in duplicate.
Okra extract stabilized o/w emulsions were centrifuged at 2727ug for 5 min
(Centrifuge 5702, Eppendorf, Hamburg, Germany) in order to separate the dispersed phase
(oil droplets) from the continuous phase and serum was then carefully collected using a
syringe. Interfacial protein concentration (*, mg m-2) was calculated as the protein
concentration difference in the extract and serum solutions divided by the specific surface
area of the oil droplets:
mg of adsorbed protein
*= (1)
SSA u mL of oil in emulsion
where specific surface area (SSA), m2/mL was calculated according to 14:
6U
SSA = (2)
d 32
where U is density of hexadecane (773 mg mL-1). Protein was measured in both solutions
and serum of centrifuged emulsions according to the Bradford method 15 using Quick
Start Bradford Protein Assay kit. Calibration curve was constructed using bovine serum
albumin (BSA) and absorption was measured at 595 nm. All measurements were
performed at least six times.
12/04/2014 12:05:31.
The isolation protocol resulted in samples with low protein content (2 and 3 % w/v for
OE4 and OE6, respectively) whereas the common structural features of pectins extracted
with the isolation protocol that was followed have been previously discussed in detail. 12
Size exclusion chromatography revealed multimodal molecular weight distribution with
three populations for both samples. The first peak to elute corresponded to polymers of
molecular weight above 1400x103 g mol-1. Two more peaks followed, corresponding to
polymers of molecular weight of 50x103 g mol-1 and 1x103 g mol-1, respectively.
The capacity of the extracts to act as emulsifiers was tested by means of emulsifying n-
hexadecane into an aqueous medium buffered at pH 3.0 containing 1.8% w/v of extract so
as to yield emulsions of = 0.2 and a nominal extract concentration in the entire emulsion
volume of 1.5% w/v. In order to quantify the capacity of these extracts towards long-term
emulsion stability, the droplet size distribution and the average droplet sizes were
measured at set time intervals. Both extracts showed good emulsification ability, producing
emulsions of d32 ~ 1213 m in the case of OE4 and d32 ~ 11 m in the case of OE6 (Table
1) both of monomodal droplet size distribution (Fig. 1).
View Online
Figure 1: Particle size distributions of emulsions prepared using 1.5% OE4 (a) and 1.5%
OE6 (b). The time development of the distributions is also shown.
12/04/2014 12:05:31.
Table 1: Influence of okra extract type and storage time (days) on the average droplet
diameters (d32 and d43) in emulsions formed with 1.5% (w/v) OE4 and OE6.a
It is commonly argued that the main contribution of such extracts towards emulsion
stability is related to the presence of co-extracted proteins that adsorb at the o/w interface
of the droplets. However, since the o/w interface is composed of a complex mixture of
proteins and polysaccharides, contribution from functional groups of the pectin structure
such as acetyl and methyl or of smaller molecules (ferulic acid moieties) can result to a
greater hydrophobicity of okra extracts, increasing their surface activity. 4 As a first step in
addressing the above, the amount of protein adsorbed at the water n-hexadecane interface
was determined both in terms of absolute value, as well as a percentage of the total protein
present in each extract. In the case of OE4, protein surface coverage of the emulsions is 0.6
mg m-2, amounting of a total 33% of the total protein present. In the case of OE6 surface
coverage is almost twice that value (1.0 mg m-2), a value close to other surface-active
protein material, while a far larger proportion (57%) of the total protein was transferred
from the bulk to the interface. The higher protein coverage can account for the differences
observed in the emulsification capacity of the two extracts although structural differences
between the samples cannot be ruled out.
The rheology of fresh okra extract solutions (continuous phase viscosity) was investigated
first. At low shear rates the viscosity of 1.8% w/v OE6 solutions was about two times
greater as compared to OE4. Although polymer components of OE6 and OE4 are of
12/04/2014 12:05:31.
similar molecular mass, their relative concentrations in the extract are different, OE6 being
richer in high-molecular weight components. The differences in viscosity of its solutions
should be attributed to the effect of the impact of the higher molecular weight of the OE6
components, but also to conformational differences of their composing macromolecules, as
has been also previously reported. 8 Concerning the rheological behaviour of emulsions
containing OE4 at different storage times the viscosity of fresh emulsions was found to be
significantly greater than that of the equivalent solutions throughout the shear rate range
(Fig. 2a). This was expected, as the presence of active fillers such as oil droplets raises the
overall viscosity of a system. Fresh OE6 emulsions demonstrated higher low-shear
viscosity than those containing OE4, i.e. ~100 mPas for OE6 and ~60 mPas for OE4
emulsions (Fig. 2). This is consistent with the viscosity data of OE4 and OE6 solutions,
which have indicated a greater low-shear viscosity for the OE6 solutions. In addition, the
OE4 and OE6 emulsions have comparable high shear rate viscosity, but at low shear rates,
the viscosity of OE6 is two orders of magnitude higher than that of OE4. This should be
attributed to the higher component of high molecular weight polysaccharides in OE6 as
compared to OE4 suggesting that at least OE6 is involved to some sort of reversible
flocculation process. This weak flocculation can originate either from weak depletion due
to unabsorbed macromolecules from the extract, or from bridging of droplets due to
interactions of the interfacial macromolecules. The tendency for self-association noticed in
the solutions could well also apply to the hydrocolloids adsorbed at adjacent droplets,
leading to a weakly flocculated droplet network. Fig. 3b presents the rheological behaviour
of emulsions containing OE6. Similarly to the previous samples, viscosity of fresh
emulsions was found to be significantly greater compared to the OE6 solutions through the
examined low shear rates.
View Online
Figure 2: (a) Influence of ageing (0-30 days) on the rheological behavior of 1.5% w/v
emulsion containing OE4 and rheological behavior of fresh 1.8% w/v okra extract solution
pH 4.0 (25 qC). (b) Influence of ageing (0-30 days) on the rheological behavior of 1.5%
w/v emulsion containing OE6 and rheological behavior of fresh 1.8% w/v okra extract
solution pH 6.0 (25 qC).
By comparing viscosities of fresh, 5 and 30 days old emulsions, it was noticed that the
latter demonstrate a significant increase in viscosity in the low shear rate regime (<1 s-1).
The effect is far more pronounced in the case of OE4-stabilized emulsions (Fig 2a). This
increase in viscosity should not be related to existing flocculated structures, as such would
tend to rearrange as to produce more compact flocs, bringing about a reduction in
viscosity. In the present case, the most plausible explanation is desorption of adsorbed
polysaccharide and protein from the interface due to the reduction in free surface area, as
the average droplet increases. This is an indication that Ostwald ripening mechanisms
should play an important role on the destabilisation of the samples. Such fresh bulk
macromolecules would both increase the bulk phase viscosity itself, and also enhance the
depletion potential, strengthening any depletion flocculation. In that rationale, the
rheological data are in complete agreement with the results of other measurements, such as
alterations in the volume-weighted average diameters (d43) and particle size distributions
between day 5 and 30, especially in OE4. It was also observed that high-shear rate
viscosity of emulsions during aging was comparable for both extracts (56 mPas),
suggesting that the floc structures are almost disrupted at high shear rates and viscosity in
such stresses depends on the rheology of the continuous phase.
View Online
4 CONCLUSIONS
5 REFERENCES
6. Sengkhamparn, N.; Sagis, L. M. C.; de Vries, R.; Schols, H. A.; Sajjaanantakul, T.;
Voragen, A. G. J., Food Hydrocolloids 2010, 24, 35-41.
7. Georgiadis, N.; Ritzoulis, C.; Sioura, G.; Kornezou, P.; Vasiliadou, C.; Tsioptsias, C.,
Food Hydrocolloids 2011, 25, 991-999.
8. Kontogiorgos, V.; Margelou, I.; Georgiadis, N.; Ritzoulis, C., Food Hydrocolloids
2012, 29, 356-362.
9. Kravtchenko, T. P.; Voragen, A. G. J.; Pilnik, W., Carbohydrate Polymers 1992, 18,
17-25.
10. Levigne, S.; Ralet, M.-C.; Thibault, J.-F., Carbohydrate Polymers 2002, 49, 145-153.
11. Thibault, J.-F., Carbohydrate Polymers 1988, 8, 209223.
12. Sengkhamparn, N.; Verhoef, R.; Schols, H. A.; Sajjaanantakul, T.; Voragen, A. G. J.,
Carbohydrate Research 2009, 344, 1824-1832.
13. Sengkhamparn, N.; Bakx, E. J.; Verhoef, R.; Schols, H. A.; Sajjaanantakul, T.;
Voragen, A. G. J., Carbohydrate Research 2009, 344, 1842-1851.
14. Walstra, P., Formation of emulsions. In Encyclopedia of emulsion technology: basic
theory, Becher, P., Ed. Marcel Dekker: New York, 1983; Vol. 1, pp 1-725.
15. Bradford, M. M., Analytical Biochemistry 1976, 72, 248254.
FUNCTIONAL PROPERTIES OF HYDROPHOBICALLY MODIFIED INULIN
Centre for Water Soluble Polymers, Glyndwr University, Plas Coch, Mold Road,
Wrexham, LL11 2AW U.K.
1 INTRODUCTION
and for its beneficial effects as a dietary fibre 3-5. It consists of linear chains of E (2,1)
fructose units with a glucose residue attached at the reducing end and has a degree of
polymerisation typically of 2 60 depending on the source. A number of researchers have
shown that inulin can be readily derivatised to produce novel surfactants that can be used
for the stabilisation of oil in water emulsions and for the encapsulation of active
compounds 6-8. Modification has mainly involved the use of organic solvents, however,
Morros et al 9,10, have recently reported modification of inulin using alkenyl succinic
anhydrides (ASA) under aqueous conditions. This reaction has been extensively used to
produce succinylated starches which have application as emulsifiers for flavour oil-in-
water emulsions in the beverages 11,12. This paper reports on work undertaken on the
synthesis, characterisation and functional properties of inulin samples chemically modified
using both octenyl and dodecenyl succinic anhydrides under aqueous conditions.
2.1 Materials
Inulin coded Fibruline DS2 (Degree of Polymerisation, DP > 10) was supplied by
Cosucra and was dried at 70 C for 24 hours before use. 2-octen-1-yl-succinic anhydride
(OSA) and 2-dodecen-1-yl-succinic anhydride (DDSA) were obtained from Aldrich
Chemical Co. and were used as received. DMSO-d6 (99.9 atom % D) was obtained from
Sigma-Aldrich Chemie GmbH. Sudan IV, a water insoluble diazo dye, was obtained from
Eastman Kodak Company and Medium Chain Triglyceride (MCT) gold oil was obtained
from Trec Nutrition UK and was used as supplied.
View Online
2.2 Methods
was made immediately after emulsion preparation and further measurements were made
periodically for up to 21 days. Before measuring the samples, background readings for the
instrument were carried out to subtract the ambient light signals from the total scattering
received from samples. Two or three drops of the sample were introduced into the
dispersion unit containing distilled water. The dispersion unit pump speed was 2000 rpm.
The obscuration was between 10% and 30%. The refractive index of the dispersing
medium and the dispersed particles were 1.33 and 1.45 respectively. Measurements were
performed in duplicate and the average values taken.
A typical 1H NMR spectrum for the modified inulin is presented in Figure 2. The prominent
peaks at 2.54 ppm and 3.34 ppm are from the solvent i.e. DMSO and HDO respectively and the
peaks from 3.14 ppm to 5.16 ppm are from the inulin. 1H NMR signals at 0.85 ppm, 1.26 ppm and
1.94 ppm correspond to the methyl and methylene groups of DDSA which indicate derivatisation
of the inulin has taken place 14,15. The extent of substitution of the ASA-inulins was calculated
from the ratio of the area of the peak at 0.85 ppm from methyl groups of ASA to the area of the
peaks from 3.14-5.16 ppm from inulin. The results are reported in Table 1 together with the
reaction efficiencies.
12/04/2014 12:05:33.
3.2 Solubility
The OSA modified inulin samples dissolved readily in water and formed optically clear
solutions at room temperature. However, the samples modified with DDSA formed slightly
cloudy solutions at room temperature but became clear on heating to 50C and remained
clear on subsequent cooling (Figure 3).
Figure 4 Dye solubilisation: volume of Sudan IV, 10mg into DDSA (12%)-inulin
0.16
DS2 native inulin
0.14 OSA(12 %)+DS2
DDSA(12 %)DS2
0.12
0.10
ABS
0.08
12/04/2014 12:05:33.
0.06
0.04
0.02
0.00
1E-3 0.01 0.1 1
Concentration / %
Figure 5 Concentration dependence of the UV-vis absorbance of native, OSA(12%)- and
DDSA(12%)- inulins in presence of excess Sudan IV.
3.4 Emulsification Properties
The droplet size for emulsions prepared using ASA-inulins are presented in Figures 6 (a)
and (b). The initial droplet sizes (surface weighted mean diameter, (d 3,2) for emulsions
prepared using OSA- and DDSA- inulin were 6.3 and 2.2 m respectively indicating the
DDSA- inulin was a more effective emulsifier. The values for both modified inulins
remained relatively constant over 21days at room temperature.
View Online
(a)
15
2% OSA(12%) at RT
10
m
d 3,2
0
initial day 3 days 7 days 14 days 21 days
(b)
15
2% DDSA(12%) at RT
12/04/2014 12:05:33.
10
m
d 3,2
0
initial day 3 days 7 days 14 days 21 days
Figure 6 Change in surface weighted mean diameter (d 3, 2) of oil in water emulsion using
(a) OSA (12%) and (b) DDSA(12%) over time.
View Online
4 CONCLUSIONS
It has been shown that inulin can be chemically modifed using alkenyl succinic anhydrides
under mild conditions in aqueous solution. Dye solubilisation studies have shown that the
derivatives have typical surfactant-like behaviour and form micellar aggregates above a
critical concentration. The CACs of OSA and DDSA modified inulins were around 0.8%
and 0.04% respectively. The fact that these materials form micelles indicates that they
have potential application as delivery vehicles for hydrophobic active compounds in foods,
neutraceuticals, pharmaceuticals, cosmetics etc. The derivatives were also found to be
effective emulsifiers and were able to form stable oil-in-water emulsions. The DDSA-
inulin produced droplets with a smaller droplet size than OSA-inulin and the droplet size
remained constant over a period of 21 days for both samples.
Acknowledgements
The work undertaken was supported through the BBSRC Integrated Biorefining Research
and Technology Club initiative [BB/I005315/1]. We are grateful to colleagues at Bangor
University who carried out the NMR measurements.
References
12/04/2014 12:05:33.
1. Meyer, D. In Handbook of hydrocolloids, 2nd ed.; Phillips, G.O., Williams, P.A., Eds.;
Woodhead Publishing Ltd: Cambridge, UK, 2009; pp 829-848.
2. Peters, D. Raw Materials in Advances in Biochemical Engineering/Biotechnology,
Eds., Ulber, R and Sell, D. Springer, Berlin, 2007, Vol. 105, Chapter 1, pp 1-30.
3. Bot, A.; Erle, U.; Vreeker, R.; Agterof, G.M. Food Hydrocolloids 2004, 18, 547-556.
4. Glibowski, P.; Pikus, S. Carbohydr. Polym. 2011, 83, 635-639.
5. Meyer, D.; Bayarri, S.; Tarrega, A.; Costell, E. Food Hydrocolloids 2011, 25, 1881-
1890.
6. Stevens, C.V.; Meriggi, A.; Booten, K. Biomacromolecules 2001, 2, 1-16.
7. Stevens, C.V.; Meriggi, A.; Peristeropoulou, M.; Christov, P.P.; Booten, K.; Levecke,
B.; Vandamme, A.; Pittevils, N; Tadros, T. F. Biomacromolecules 2001, 2, 21256-1259.
8. Exerowa, D.; Gotchev, G.; Kolarov, T.; Kristov, K.; Levecke, B.; Tadros, T. F. Colloids
Surf., A 2009, 334, 87-91.
9. Morros, J.; Levecke, B.; Rosa I.M. Carbohydr. Polym. 2010, 81, 681-686.
10. Morros, J.; Levecke, B.; Rosa I.M. Carbohydr. Polym. 2011, 84, 1110-1116.
11.Wurzburg, O. T., In Food Polysaccharides and their applications, 2nd ed.; Stephens,
A.M., Phillips, G.O., Williams, P.A., Eds.; Taylor and Francis group, Florida, USA , 2006,
p 87-118.
12. Bai, Y.; Shi, Y.C. Carbohydr. Polym. 2011, 83, 520-527.
13. Kokubun, S.; Ratcliffe, I.; Williams, P.A. Biomacromolecules, 2013, 14, 2830-2836.
14. Chi, H.; Xu, K.; Xue, D.; Song, C.; Zhang, W.; Wang, P. Food Res. Int. 2007, 40, 232-
238.
15. Tizzotti, M.J.; Sweedman, M.C.; Tang. D.; Schaefer, C.; Gilbert, G. R. J. Agric. Food
Chem. 2011, 59, 6913-6919.
STABILISATION OF FOAMS BY WHEY PROTEIN GEL PARTICLES
1 INTRODUCTION
There is a rising trend of consuming gourmet beverages, where a large proportion of the
final structure is milk foam. At the same time, like most fast moving consumer goods,
convenience is driving the market, making it important that the consumer gets the
experience of a speciality beverage with as little effort as possible. This encompasses
technical and scientific challenges in order to deliver foam on a beverage with acceptable
taste and texture and that the structure can be produced without the need of special devices
12/04/2014 12:05:35.
or particular skill.
Foams are thermodynamically unstable systems and generally have a lifetime of some
orders of magnitude smaller than that of emulsions (hours compared to months). The main
reasons for this are that the films between bubbles are generally much larger than between
emulsion droplets and the interfacial tension on the air-water interface is larger than on the
oil-water interface by a factor of around 5. Also, bubbles are typically larger and less dense
than oil droplets and therefore gravity creaming is much faster for foams than for o/w
emulsions1. Foams are therefore prone to drainage, disproportionation (the equivalent of
Ostwald ripening in emulsions) and coalescence2. Additionally, when considering foams,
the different instability mechanisms can reinforce each other by acting synergistically
towards the collapse of the foam. For example, disproportionation results in larger bubbles
and therefore larger films which in turn accelerate drainage, continuously decreasing the
foam stability3.
The parameters that most affect the characteristics of foams are the viscosity of the
liquid, the interfacial properties of the adsorbed layer and the disjoining pressure between
adsorbed layers in liquid thin films 4.
Following the need to increase the stability of foams while trying not to introduce new
ingredients to formulations (because of the trend towards clean labels), the necessity of
giving new properties using the existing components in milk becomes significant. The
whey protein fraction of milk has been widely used in food formulations because of its
high nutritional value but also the broad functional properties it encompasses5. Whey
protein isolates (WPI) are products with protein content larger than 90% and have been
extensively studied in terms of their ability to stabilise foams6. The effect of heat treatment
on the structure and functional properties of whey proteins has also been the subject of
research 7,8. The effect of the heat induced aggregation of whey proteins on their foaming
properties has been investigated and it was shown that polymerisation of whey protein
fractions by sulfhydryl-disulfide interactions improve the foaming properties9. While
View Online
proteins can form films with high interfacial elasticity and viscosity through the various
forms of cross-linking of the adsorbed molecules on the interface, it was found that they
still cant completely stop bubble shrinking and foam instability mostly due to their
tendency to desorb to the bulk10.
Another solution of stabilising air bubbles is by solid particles that have an affinity for
the air water interface and can create a network around the bubbles and stabilise them.
Solid particles will act as antifoam unless they are of the correct size and have an optimum
contact angle with the interface they are adsorbed to. It is in fact the synergistic effect of
particles along with surfactants that can lead to altering the interfacial properties of the
particles and produce very stable foams 11.
This study shows the effect of particles obtained by thermal gelling of whey proteins
in the stability of foams. Thermal gelation of proteins is a two step mechanism where the
first step involves the denaturation and unfolding of the whey protein molecules followed
by the rearrangement and final aggregation of the functional groups that become exposed
forming a three dimensional network5. The heat induced denaturation and gelling
properties of whey proteins is used as the mechanism of obtaining protein particles which
are acquired after drying and reducing the size of whey protein gels. The dried powders
obtained, upon reconstitution can produce cold-set weak gel structures12. Since protein
unfolding and aggregation is sensitive to pH and ionic strength5, the properties of the dried
gel particles are also prone to these factors. The positive effect of these dried gel particles
on foam drainage has been demonstrated in the past13. The aim of this study is to provide
evidence of the mechanism in which these particles can produce films with enhanced
mechanical properties that can withstand instability and relate this to the rheological
12/04/2014 12:05:35.
2 METHODS
2.1 Materials
Whey protein isolate (BiPRO) was obtained from Davisco Foods International Inc. (Le
Sueur, MN). Protein solutions were prepared in double distilled deionised water and the
pH was adjusted using NaOH or HCl (5 M) and NaN3 (0.001 wt%)was added to prevent
microbiological growth. All chemicals were obtained from Sigma Aldrich (Dorset, UK).
From this point and on the pH values of the particles refer to the pH at which the gel
particles were manufactured.
dispersion unit attached (Malvern Instruments, UK). The scattered intensity as a function
of the angle was transformed into size distribution using the Mie theory and the relevant
refractive index for the protein (1.456) was used.
2.2.4 Rheological characterisation of the bulk solutions and the resulting foams
Viscosity and rheological measurements were performed using a Kinexus rheometer by
Malvern Instruments (Malvern, UK). Flow curves of the protein solutions were obtained
by using a cup and bob geometry (25 mm and 27.5mm diameter respectively) and applying
a range of shear rates (0.1 100 s-1) while measuring the viscosity. The flow curves of
foams were determined using a 60 mm serrated parallel plate geometry and applying a
smaller range of shear rates (0.1 10 s-1). The rheological properties of both the bulk phase
and the foams were determined by oscillatory rheology by applying a strain controlled
frequency sweep at a strain rate within the linear viscoelastic region of the samples as
defined by an amplitude sweep performed beforehand using a similar sample. In the case
of the bulk phase the oscillations were performed using the 60 mm sandblasted parallel
plates and in the case of the foams the 60 mm serrated plates in order to eliminate slippage.
water was also assessed and compared to the effect of WPI. The results show that the
surface tension lowering effect was stronger for the non-thermally treated WPI proteins as
expected (Figure 2d). The solutions containing the particles prepared at pH 5 also had
similar surface tension values (Figure 2a) to the non-treated WPI probably due to the fact
that after gelling at this pH there are still significant amounts of non-aggregated proteins.
The suspensions containing the particles made at pH 7 and 8 show similar surface tension
values that are higher than the control and the suspensions of the particles gelled at pH 5
(Figure 2b and c). Additionally, the suspensions with the pH 7 and 8 particles show greater
dependency on concentration with a concentration of 3% having lower surface tension than
1% and 5% which is counter intuitive.
6
Volume %
0
1 10 100 1000
60 60
58 58
surface tension (mN/m)
54 54
52 52
50 50
48 48
46 46
0 100 200 300 400 500 600 0 100 200 300 400 500 600
62 62
60 60
58 58
surface tension (mN/m)
54 54
52 52
50 50
48 48
46 46
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Figure 2: Surface tension over time curves of gel particle suspensions prepared at pH 5
(a), pH 7 (b), pH 8 (c) and native WPI solutions (d) at concentrations of 1% (), 3% ()
and 5% ().
12/04/2014 12:05:35.
1 1
viscosity (Pa s)
viscosity (Pa s)
0.1 0.1
0.01 0.01
0.001 0.001
0.0001 0.0001
0.00001 0.00001
0.1 1 10 100 0.1 1 10 100
10 10
1 1
viscosity (Pa s)
viscosity (Pa s)
0.1 0.1
0.01 0.01
0.001 0.001
0.0001 0.0001
0.00001 0.00001
0.1 1 10 100 0.1 1 10 100
c) d)
shear rate (1/s) shear rate (1/s)
Figure 3: Flow curves of gel particle suspensions prepared at pH 5 (a), pH 7 (b), pH 8 (c)
and native WPI solutions (d) at concentrations of 1% (), 3% () and 5% ().
12/04/2014 12:05:35.
1
G' (Pa) at 1 Hz
0.1
0.01
0.001
0 1 2 3 4 5 6
% wt
Figure 4: Elastic modulus (G) at 1 Hz of the dispersions of gel particles made at pH 5
(), pH 7 () and pH 8 () and the WPI solutions () at the 3 different concentrations.
particle suspensions do not seem to significantly affect the bulk rheological properties of
the suspensions.
Figure 5: Viscosities of foams at 0.25 s-1 stabilised by gel particles made at pH 5 (), pH 7
() and pH 8 () and WPI () at the 3 different concentrations.
All the foams displayed shear thinning behaviour which can be explained by the disruption
of the bubble structure under shear (data not shown). In terms of viscosity values the
viscosities at 0.25 s-1 are presented in Figure 5 to facilitate comparison. The viscosity of
the foams made with gel particle suspensions made at pH 8 was higher than the rest and
not significantly dependent on the concentration. At a concentration of 5% the foams
stabilised by gel particles gelled at pH 7 were equally high to the ones at pH 8 but showed
an increase with increasing concentration. Finally, the foams stabilised by the gel particles
made at pH 5 and the WPI showed the lowest viscosity values at all concentrations. It
12/04/2014 12:05:35.
seems therefore that the gel particles made at pH 7 and more profoundly at pH 8 are able to
create foams more resistant to deformation than the foams stabilised with gel particles
made at pH 5 and non-treated WPI.
The stability of foams stabilised with the reconstituted dried gel particles was investigated
in terms of foam half-life. The volume of foams created by air sparging was monitored
over time and the time corresponding to the reduction of the foam to half of the original
volume was recorded.
The results (Figure 6) show a significantly increased stability of the foams stabilised
by the gel particles created at pH 8. The increase in stability has strong correlation with the
increase of concentration within the range studied. Gel particles made at pH 5 also showed
a moderate but significant increase in foam stability that was not concentration dependent.
This indicates that the conditions under which the whey protein gels are made have a
strong impact on the functionality of the dried particles obtained from them in terms of
foaming
1000
600
400
200
0
1% 3% 5%
. protein %wt
Figure 6: Foam stability plot showing the half-life in minutes of the gel particle
suspensions prepared at pH 5 (), pH 7 (), pH 8 () and native WPI solutions () at the
different concentrations.
12/04/2014 12:05:35.
Figure 7 Confocal micrographs of foams stabilised with reconstituted gel protein particles
gelled at pH 5 (a), pH 7 (b) and pH 8 (c).
have an affinity to the air water interface as it is revealed by their presence only on the
surface of the air bubbles. In the case of the gel protein particles prepared at pH 7 and 8 the
particles are larger and irregular in size. In this case the affinity to the air water interface is
also present and the particles seem to block completely the plateau borders between the
bubbles. This compact structure is probably the main reason why the particles made at pH
8 have higher performance in terms of foam stability, allowing less drainage and blocking
the path of liquid from flowing between the bubbles.
4 CONCLUSIONS
more resistant to flow. These particles as shown in the micrographs have the tendency to
align on the air water interface and create physical barriers which along with the
electrostatic repulsion but also the possible water binding capabilities can significantly
reduce the factors that cause foam instability.
While both gel particles created at pH 7 and 8 upon reconstitution in water have a
similar viscosity increasing effect they dont show any significant effect on the mechanical
properties of the bulk solution in terms of storage or loss modulus. Furthermore these two
different preparations (pH 7 and 8) while they have several similarities when in the bulk
suspensions in terms of particle size, surface tension and rheological properties they show
different performance when foamed. The reason of this effect needs to be further
investigated in order to understand completely the mechanism with which the pH 8 gel
particles stabilise liquid foams.
Finally the particles created by gelling close to the isoelectric point of whey
proteins at pH 5 exhibit similar effects in terms of rheological and interfacial properties
with the whey protein isolate suspensions probably due to their small size and lack of
electrostatic charge. Their foam stabilisation contribution was higher than the control but
still significantly lower than the particles made at pH 8.
References
1. E. Dickinson, Current Opinion in Colloid & Interface Science, 2010, 15, 4049.
2. J. E. Kinsella, Food Chemistry, 1981, 7, 273288.
3. P. Walstra, Physical chemistry of foods, CRC Press, 2003.
4. P. A. Wierenga and H. Gruppen, Current Opinion in Colloid & Interface Science, 2010,
12/04/2014 12:05:35.
15, 365373.
5. D. M. Mulvihill and M. Donovan, Irish Journal of Food Science and Technology, 1987,
11, 4375.
6. E. A. Foegeding, P. J. Luck, and J. P. Davis, Food Hydrocolloids, 2006, 20, 284292.
7. J. N. deWit and G. Klarenbeek, Journal of Dairy Science, 1984, 67, 27012710.
8. M. e Mangino, Y. y Liao, N. j Harper, C. v Morr, and J. g Zadow, Journal of Food
Science, 1987, 52, 15221524.
9. H. Zhu and S. Damodaran, J. Agric. Food Chem., 1994, 42, 846855.
10. B. S. Murray and R. Ettelaie, Current Opinion in Colloid & Interface Science, 2004, 9,
314320.
11. B. P. Binks, M. Kirkland, and J. A. Rodrigues, Soft Matter, 2008, 4, 2373.
12. H. M. Hudson, C. R. Daubert, and E. A. Foegeding, J. Agric. Food Chem., 2000, 48,
31123119.
13. J. D. Firebaugh and C. R. Daubert, International Journal of Food Properties, 2005, 8,
243253.
14. R. D. Waniska and J. E. Kinsella, Journal of Food Science, 1979, 44, 13981402.
15. C. Schmitt, C. Bovay, M. Rouvet, S. Shojaei-Rami, and E. Kolodziejczyk, Langmuir,
2007, 23, 41554166.
16. J. L. Klemaszewski and J. E. Kinsella, J. Agric. Food Chem., 1991, 39, 10331036.
17. M. Verheul and S. P. F. M. Roefs, Food Hydrocolloids, 1998, 12, 1724.
ETHOCEL FOR OIL STRUCTURING IN FOOD APPLICATIONS
1
Dow Pharma and Food Solutions, August-Wolff-Str.13, 29699, Bomlitz, Germany
2
Dow Pharma and Food Solutions, Larkin Laboratories, Midland MI 48674, USA
12/04/2014 12:05:37.
Consumption of saturated fat is linked with the increase of cardiovascular disease, while
the consumption of monounsaturated and polyunsaturated fats decrease these risks. 1
Saturated fats have beneficial attributes in food formulations as the solid structure and
texture of fats is provided by saturated triacylglycerols. Therefore, it is desirable to develop
healthier alternatives by creating structure in edible oils.
In edible oils, ethyl cellulose is capable of forming a gel when the solution is first heated
above the glass transition temperature. Upon subsequent cooling, the ethyl cellulose forms
a polymer network stabilized by physical interactions entrapping the liquid-phase oil
within.2 Due to the restricted mobility and migration of the oil inside the polymer network,
ethyl cellulose structured healthy oils may potentially provide the solid-like properties of
saturated triacylglycerols without the high levels of saturated fatty acids.
The replacement of saturated and trans fats by healthy oils while retaining structural
properties of solid fats is highly desirable for various food applications e.g. in emulsions,
bakery products and meat products. Zetzl et al have investigated recently that cooked
frankfurters made with ethyl cellulose structured healthy oils showed no significant
differences in chewiness or hardness compared to the control products made with beef fat.3
We were able to reduce saturated and trans fat in a broad range of bakery products. The
structuring ability and mouth feeling of saturated fats is especially important in laminating
View Online
pastry due its high fat content. The saturated fat content was reduced by > 30% in Danish
Pastry, Puff Pastry as well as Croissant by substitution of butter or margarine by ethyl
cellulose structured healthy oils. On top we could show that it is possible to reduce the
total amount of fat in the final bakery products.
REFERENCES
1 R.P. Mensink, P.L. Zock, A.D.M. Kester and M.B. Katan, Am. J. Clin. Nutr., 2003,
77, 1146.
2 A.J. Gravelle, S. Barbut, A. G. Marangoni, Food Res. Int., 2012, 48, 578.
3 A.K. Zetzl, A.G. Marangoni and S. Barbut, Food Funct., 2012, 3, 327.
12/04/2014 12:05:37.
USE OF POLYSACCHARIDES AS STABILISERS FOR SPECIALIZED OXYGEN
COCKTAILS
Saratov State Agrarian University named after N.I. Vavilov, Saratov, Russia
n.ptichkina@gmail.com
1 INTRODUCTION
the internal organs 1. The main component of the oxygen cocktail is the foaming and
stabilizing component which provides the formation of foam in the beverage. The stability
of foams depends on the nature and concentration of the foaming agent, the dispersion
properties of the medium, temperature, mechanical stress, and etc. Stabilization of foams is
provided by surfactants2-7.
The aim of the present work was to establish the process for the production of
oxygen cocktails based on liquid whey, polysaccharides and natural juices with the
replacement of the traditional foaming reagents, i.e syrup of licorice root, egg white and
gelatin8,9.
Using polysaccharides and whey proteins as stabilizers in the production of oxygen
cocktails is a novel area of work. It is expected that this approach will advance existing
formulations of oxygen beverages with improved functional properties, sensory and
structural characteristics and reduced caloric impact10,11.
2.1 Materials
The materials utilized in the study were gelatin (Kazan Gelatin Factory, Russia),
high methoxy pectin (LMP), locust been gum (LBG) and guar gum with a molecular
weight of 30, 100, 400 kDa (Danisco, France).
View Online
In order to obtain a liquid whey in vitro the full cream milk was pasteurized at 72-
75 0C for 15-20 s. Then the system was cooled down to 30-32 0C and mesophilic cultures
(S. Lactis, S.Cremoris) were added, and allowed to ferment for 8 h at 30-32C until the pH
was reduced to 4.0-4.2. The system was stirred at a temperature of 45-50 0C, and liquid
whey was separated off. It was found to contain 1.2% of protein, 0.3% of fat and 4.7% of
lactose.
the beverage was produced using medical oxygen (99.9% of pure medical grade oxygen)
from the oxygen cylinder. Tested samples were poured into a graduated cylinder at 2-4 C,
followed by churning and bubbling with medical oxygen from a mixer to achieve the
desirable foam properties. Oxygen was sparged into the system at a rate between 0.5-5
liters / min.
2.3 Methods
The amount of total protein in liquid whey was established using a total
nitrogen/protein analyzer (Rapid N cube, Germany). Amino acid composition for liquid
whey was determined using an amino acid analyzer (Aracus, Germany) and using liquid
chromatography on the Varian ProStar 500 Series (Canada). The concentration of
lactose in the obtained whey was determined using a refractometer (IRF-464, Russia). The
amount of fat in the liquid whey was established by the Babcock method. pH of the liquid
whey-polysaccharide dispersions was determined on the Checer ( Hanna, USA).
Foaming ability was determined as the ratio between the quantity of dispersion
medium and the dispersed phase (equation 1):
where Vf volume of foam, ml; Vl - volume of liquid before whipping process, ml.
View Online
Foam stability was established by measuring the strength and lifetime of the foam
within 3 min which is required for assessing the foam as stable12.
3.1 Justification of the choice of cottage cheese whey as the basis of oxygen cocktails
3.2 Study of the properties of oxygen foams depending on the nature of the foaming
agents
Outcomes from Table 3 revealed that the commonly used foaming agents were able
to provide foam with sufficiently high multiplicity compared to whey protein. Foam
stability for all samples was low, i.e. within 2 min, where the foams indicate a rapid loss in
their structure, and eventually collapsed.
3.3 Application of polysaccharides for the stabilization of the protein oxygen foam
The systems with 0.8% of whey protein with the addition of fruit juices and
polysaccharide have been prepared according to the experimental procedure in the
Materials and Methods section. For this, the dairy systems with HMP, LBG and guar gum
with molecular weight of 30, 100 and 400 kDa were analyzed for foaming ability and
stability.
View Online
As shown in Figure 1, foams with 0.1-0.5% of HMP, 0.1-0.2% of locust been gum,
0.1-0.3% of guar gum, had greater values of foaming ability. Increasing the concentration
of polysaccharides up to 0.5% results in systems with high viscosity, and therefore foam
ability is reduced. In foam preparations with low concentration of polysaccharides air
bubbles with large size are dominant in the system, whereas with increasing the viscosity,
as in the case of foams with higher amount of stabilizers, air bubbles become smaller. As a
result, the foam resists whipping and values of foaming ability were reduced.
Moreover, increasing the molecular weight of guar gum leads to greater values of
viscosity, and, as a consequence, dropping of foam multiplicity. To obtain a system with
good foam properties, the concentration of the stabilizer in the system should be reduced
with increasing its molecular weight.
Mass fraction of
Concentration of
Injection amino acids in
Amino acids amino acids in
volume, mcl whey, %
whey, nmol
mg/100 g
Table 3: Foam ability and stability for the foams with different stabilizing agents
During the investigation on the effect of stabilizers on foam properties, it was found
that the best temperature for the foam formation is not more than 4 0C. With increasing the
temperature up to 25 0C, foaming capacity was reduced due to thermal motion of protein
molecules that are not capable of strong adsorption on the surface of the foam within the
experimental settings.
The foam stability for the preparations containing liquid whey and fruit juice (2:1),
and polysaccharides, within 3 min was also investigated (Figure 2). Results revealed that
foams with currently utilized foaming agents, (i.e. licorice root, egg white and gelatin)
became unstable, whereas foams with polysaccharide remained stable even after one hour
after preparation.
Based on the data obtained, polysaccharides that provide consistency for the
oxygen beverages, are as follows: 0.1 to 0.3% LBG, 0.1 to 0.5% HMP, 0.1 to 0.3% guar
gum with molecular weight of 30 : D0, 0.1 to 0.2% guar gum with molecular weight of 100
and 400 kDa. Foam stability of preparations is improved up to 20 times in the presence of
polysaccharide as compared to currently used foaming agents.
View Online
450
400
Foam ability, %
350
300
250
200
150
100
50
0
0 0,1 0,2 0,3 0,4 0,5 0,6
! oncentration, %
Figure 1: Foam ability for the systems containing liquid whey and fruit juice (2:1), and
polysaccharides: HM pectin (z), guar gum 30 kDa (S), guar gum 100 kDa (),
guar gum 400 kDa (), LBG (), depending on the concentration at 4 C.
12/04/2014 12:05:38.
450
400
Volume of oxygen foams, Am3
350
300
250
200
150
100
50
0
1 2 3 4 5 6
Oxygen foams
Figure 2: The volume of oxygen foams after preparation (closed symbols) and
stability of foams with licorice root (1), egg white (2), gelatin (3), whey protein (4), whey
protein and LBG (5), whey protein and HMP (6) within 3 min.
4. CONCLUSIONS
5. ACKNOWLEDGEMENTS
The authors would like to thank Professor Peter A. Williams and Dr. Graham
Sworn for samples of guar gum.
References
F. Debeaufort
1 INTRODUCTION
For sustainable and ecological reasons, the interest of the industry for renewable, bio-
sourced and biodegradable polymers for packaging is rising significantly. Ten years ago,
less than 10 companies were able to produce and to market bio-based packaging films.
Today, more than 250 companies in Europe sell packaging materials (films, sachets, trays,
dishes, blisters, cups.) made with natural bio-polymers as shown in Figure 1. A major
defining difference between hydrocolloid-biopolymers and other synthetic plastic polymers
can be found in their structures. All polymers are made of repetitive units called
monomers. Biopolymers often have a well-defined structure, though this is not a defining
characteristic like cellulosic or starch materials: The exact chemical composition and the
sequence in which these units are arranged is called the primary structure, in the case of
proteins or chitosans. Many biopolymers spontaneously fold into characteristic compact
shapes (such as "protein folding" as well as secondary structure and tertiary structure),
which determine their biological functions and depend in a complicated way on their
primary structures. These particular structures evolve according external factors such as
temperature and relative humidity or pressure. In contrast most synthetic polymers have
much simpler and more random (or stochastic) structures.
12/04/2014 12:05:39.
The general properties of film-forming materials should include (1) good wetting of the
surface to be protected; (2) firm binding of the particles, inclusions or droplets (emulsions)
in the film; (3) rapid drying in a thin layer (from a few seconds to hours in a range of 15 to
200C), with formation of strong moisture- and gas-resistant films that can withstand
prolonged action of the external or coated medium; and (4) good adhesion to the surface
being protected. In many cases, these properties are often obtained by combining film-
forming materials with other components such as plasticizers, emulsifiers, tension-active
12/04/2014 12:05:39.
agents, fibres, reticulation agents etc, as well as by combination of two or more film-
forming biopolymers.
A huge list of hydrocolloids have been studied, characterized or patented for their film-
forming properties, but most of them contain a plasticizer (Table 1). The film-forming
ability of biopolymers/hydrocolloids is mainly assessed from mechanical properties
measured by tensile test. However, the characterisation in shape of film instead of coating
requires thicker thickness and requires addition of plasticizers which do not permit real
comparison between hydrocolloids. On the contrary, when used as coating, the support
provide mechanical resistance and film-forming ability could be assessed from scratch test
or surface properties measurement (spreadability, wettability, and cohesion coefficients).
Unfortunately, these techniques are common for paints, varnish, gold coatings etc. but not
for hydrocolloids.
View Online
Table 1: List of hydrocolloids cited as film-forming agents for edible film and coating
applications, from scientific literature, technical notes and patents.
Gum arabic
Inulin
Konjac gum
Locust bean gum
Pectin
Starches and modified
starches
Tara gum
-Glucan
12/04/2014 12:05:39.
Table 2: Physicochemical properties of some hydrocolloids as films and comparison to common synthetic polymer films (plastics): film-
forming ability (FFA), mechanical resistance (tensile strength TS, Elongation E), barrier properties (WVP, PO2, PCO2).
Film/coating composition FFA TS (MPa) E WVP PO2 P CO2
(%) (10-11 g/(m.s.Pa)) (10-10 cm3/(m.s.Pa)) (10-10 cm3/(m.s.Pa))
Polysaccharides
Methylcellulose +++ 7,55 0,7 13
Chitosans +++ 58 3,5 5,04 0,015 9,5
Methylcellulose+chitosan +++ 6,7 /
Corn starch + 32 37 36,8 2,48 41,9
Emulsions, Foams and Films
Proteins
Soy protein+Gly ++ 3,5 125 303 36,4
Corn Zein + 11 3,5 53,5 542
Fish gelatin + 25,9
Whey protein+sorbitol ++ 14 1,5 71,7 35
Whey protein+beeswax + 5 1 39,3
Pig Gelatin + 11 35 16 4
wheat gluten ++ 2 90 125 32
Caseinate Na + 37 2,5 42
Egg white + 5 62,5 230 567
From the different papers dealing with the film-forming properties, it seems that the best
hydrocolloids able to form films are cellulose derivatives, pullulans and starches and then
alginates carrageenans, but less for proteins which always require high amounts of
plasticizers7. Indeed, other physico-chemical properties of hydrocolloid based films and
coatings, particularly mechanical and barrier properties are strongly dependent on their
film forming ability (Table 2).
A new innovative way to improve and form films from charged hydrocolloids is the layer-
by-layer deposition of positively and negatively charged biopolymers. Indeed, solutions of
pectin and chitosan could be applied successively, creating ionic interactions between the
two polymers which enhance film formation, mechanical properties and consequently
barrier performances 8. Elsewhere, Bayramoglu and Stroeve 9 prepared films composed of
alternate nano-layers of lysozyme and bovine haemoglobin which provide significant
benefit in edible packaging properties compared to either lysozyme or haemoglobin films
or a mixture of both proteins.
Mechanical properties
Edible films and coatings are targeted to be eaten, and then mechanical resistance must be
low enough not to be perceived during consumption. Mechanical properties of films and
coatings based on hydrocolloids are, however, often compared to common plastic polymer
films but there is no requirement to formulate films to reach values close to that of plastic
films. For polysaccharides films, tensile strength (TS) values vary between 10 and 100
12/04/2014 12:05:39.
MPa whereas the E values range between 1 and 80%. Generally, protein films have a lower
TS while their E values are higher10. Comparison with common synthetic polymers used in
film-forming preparations, TS values approach those of polysaccharides films: 9 to 20
MPa for low-density polyethylene (LDPE) and 10 to 60 MPa for high-density polyethylene
(HDPE). However, LDPE and HDPE have higher E values (up to 1000%), which are
greater than those of protein films. Thus, it is necessary to add plasticizers into
polysaccharide or protein film-forming preparations to increase E values. However,
increasing the concentration of these plasticizers resulted in decreased TS. For example,
films produced with low molecular weight chitosan at 3% w/w in 1% acetic acid, with
glycerol as a plasticizer at 0.25 and 0.50 mL/g of chitosan, were reported to have a TS of
15 to 35 MPa and percent elongation at break (%E) of 17 to 7611.
Surface properties
Very few data are available dealing with the surface characteristics of hydrocolloid films
and coatings. Two strategies are based on contact angle measurements:
View Online
-1) determination of the contact angle and from these are deducted the works of adhesion
and cohesion that are of key importance for assessing their ability to cover the surface to
coat
-2) determination of liquid absorption flux and swelling from kinetics of wetting (angle
and droplet volume) that give an indication of the behaviour of films or coatings when they
are exposed to liquids or to high moisture content foods like gels.
Moisture barriers
Hydrocolloids are water soluble biopolymers and have poor moisture barrier properties and
very weak resistance to water absorption. Coatings/films without plasticizer often yield to
significantly higher WVP values than those with plasticizer. This difference is often
attributed to the presence of pores and cracks in unplasticized coatings, as observed by
Garca et al. 21. In the absence of cracks or pores, a plasticized film generally would be
expected to exhibit higher WVP than an unplasticized film 22. However, compared to
expected moisture barrier performances as plastic polymers (polyethylene, polypropylene,
PET), hydrocolloids are too permeable to water vapour for preservation of moisture
sensitive foods. This is the reason why hydrocolloids are often associated with
hydrophobic non-soluble components to form composite materials.
Hydrocolloids such as proteins, starches or celluloses can be combined with pure lipids
either by incorporating the lipids in the hydrocolloid film-forming solution (emulsion
technique) or by depositing the lipid layers onto the surface of the pre-formed hydrocolloid
film to obtain a bilayer 23. Multi-component films have been extensively reviewed by Wu
et al. 13 and Debeaufort et al. 24. The addition of non-lipid compounds or particles such as
sugar crystals, fibres, and proteins as dispersed components in fat materials permits
formation of fat dispersions, such as chocolates. These composite films take advantage of
the functional properties of each component of the film to provide both barrier and
mechanical properties. The resulting water barrier efficiency of bilayered films is often of
the same order of magnitude as that of pure lipid 24 and is much higher than that of
emulsion-based films 25, 26 or of pure hydrocolloid films. However, the hydrocolloid layer
is hydrophilic, and tends to absorb water when the film is in direct contact with a moist
View Online
phase. Furthermore, the additional processing steps for film making (casting and drying)
make bilayer films difficult to use in high-speed commercial production.
But some applications of hydrocolloid based coatings allow reduction of the moisture loss
by food during storage and transportation. A cheese coating to prevent moisture loss was
demonstrated by Kampf and Nussinovitch 27. Kappa-carrageenan, alginates and gellan
could be very beneficial and of commercial importance to the dairy industry in terms of
cheese weight preservation, appearance, taste, edible packaging and post-contamination
prevention.
Gases barriers
Gas barrier of hydrocolloids ranges on at least six orders of magnitude and strongly
depends on the plasticizer content as well as on the hydration level (water content or
relative humidity). Very few studies concern oxygen permeability measurements of
hydrocolloid based films at various humidities. Indeed, all values given in the Table 2 were
measured at 0% RH, and then the oxygen permeability looks very promising as oxygen
barrier. Usually, hydrocolloid films present a very high barrier to oxygen, they are mostly
permeable to water vapour. However, the impact of moisture on the oxygen and other gas
transfers is deleterious. Kurek 29 displayed an increase of the gas permeability of chitosan-
based film of 540 times and 120 times respectively for oxygen and carbon dioxide
permeabilities. In the same way, Fabra et al. 30 observed increases of oxygen and carbon
dioxide permeabilities of various caseinate-based edible films ranging from 2 up to more
12/04/2014 12:05:39.
One possible way suggested to lower interactions between aroma compounds and plastics
and responsible of flavour perception loss or unbalanced sensory, is to retain flavour
molecules inside the food product by using an additional barrier on the food surface. This
could be an edible film or coating or a thin layer having a high selectivity against aroma
transfer, and that can be eaten along with the protected food 2,31. The main application is
only using barrier properties of the hydrocolloid coatings to retain aroma within the food;
however, the film or coating could be used as a carrier or support for flavours at the surface
of the product. These volatile compounds will be released rapidly when the consumer
tastes the product. Indeed, several products are already on the market using this technology
for flavouring. For instance, there is a roasted peanut with a curry flavoured coating which
is instantaneously dissolved in the mouth and gives immediately the perception of the
Indian spice. Another example designed for children, is a multi-sugar-coated sweets in
which each layer of the coating contains different tastes and flavours separated by arabic
gum or hydrocolloid layers to prevent migration of aroma compounds from one layer to
another one. For this application, diffusivity of volatile compounds should be very low and
with a high affinity for the coating, which should be highly soluble in the mouth. As
previously outlined, edible films and coatings can deliver and maintain desirable
concentrations of colour, flavour, spiciness, sweetness, saltiness, etc. Several commercial
films, especially Japanese pullulan-based films, are available in a variety of colors, with
View Online
Various authors have shown that hydrocolloid coatings allow control of either the retention
of solutes on the food surface, or preventing solute migrations from or into the food. In the
last 20 years, many papers dealing with the release of solute by edible coating and films
have been published. For example, Vojdani and Torres 35 studied the barrier properties of
cellulose derivatives to sorbates (antimicrobial) or Giannakopoulos and Guilbert 36
displayed that protein-lipid coating, prepared from emulsions limits the diffusion of an
antifungal agent into papaya semi-dried slices and delays significantly the development of
aspergillus and other fungi. More recently, Fabra et al. 18 controlled the release in aqueous
solutions of various compounds, D-Limonene, N-hexanal (flavours), rutine (antioxidant)
and methyl-paraben (preservative) from carrageenan based films by the way of the nature
and concentration of ions (Na+, K+, Li+) added either in the liquid media or in the film
12/04/2014 12:05:39.
recipe. Other applications for nutritional benefits on fried foods mainly deal with the
reduction of oil absorption and reduction of moisture loss. Gellan gum and guar gum
applied as coatings on a food matrix affects positively the heat transfer during a frying
process and consequently the oil uptake. While the potato strips coated with the
hydrocolloid solutions were fried at 170C, the hydrocolloid coatings significantly reduced
the heat transfer coefficients as well as oil uptake which became more apparent at higher
concentrations 37. The same trend was observed for egg white coated fried chips 38.
Edible coatings provide a physical barrier against mass transport from the environment to
food and from food to the environment. If these barrier properties are important for food
passive protection, consumers ask nowadays for better food safety and for higher
nutritional and flavour properties. In recent years, active packaging was developed to
extend food shelf-life by increasing the coatings positive effect. For example, more
activity can be provided to edible coatings by adding active compounds, such as flavours,
antibacterials, antioxidants or nutraceuticals. Active compounds can be incorporated
directly into the edible hydrocolloid matrix or can be pre-encapsulated before being
incorporated in the matrix to better protect active compound activity and properties. The
Eastern Countries (particularly Japan) are the most advanced in this particular domain,
North America follows and Europe is now developing more and more active material
solutions.
Flavour compounds can be used to obtain active edible packaging since they can act as
antimicrobials, antioxidants and flavouring agents. In particular, essential oils can be added
View Online
to edible films and coatings to modify flavour, aroma, and odour, as well as to introduce
antimicrobial properties 39.
Very little published data exist on the incorporation of plant essential oils into edible films
and coatings. Essential oils are regarded as alternatives to chemical preservatives, and their
use in foods meets the demands of consumers for minimally processed natural products, as
reviewed by Burt 40. Vanillin has been used recently as a bacteriostatic rather than a
bactericidal agent in fresh-cut apples 41. Essential oils have also been evaluated for their
ability to protect food against pathogenic bacteria in contaminated apple juice 42,43 and
other foods and they are used as flavoring agents in baked goods, sweets, ice cream,
beverages, and chewing gum 40,44. Sanchez-Gonzalez et al. 45,46 introduced tea tree and
bergamot essential oils in chitosan or hydroxypropylmethylcellulose edible films at a range
of 0-3% (w/w) in the film forming suspension for antimicrobial properties. They displayed
tantimicrobial efficiency at the higher concentration of bergamot on Penicillium italicum
and at the lowest rate of tea tree oil on Listeria monocytogenes.
Rojas-Grau et al. 47 recently investigated the effect of plant essential oils on antimicrobial
and physical properties of apple puree edible films. Alginate-apple puree films, containing
plant essential oils, were further explored as edible coatings by Rojas-Grau et al. 48 with
the aim of studying the effect of lemongrass, oregano oil and vanillin on native
psychrophilic aerobic bacteria, yeasts, moulds and inoculated Listeria innocua in fresh-cut
Fuji apples. Coatings with essential oils seemed to effectively inhibit the growth of L.
innocua inoculated on apple pieces as well as psychrophilic aerobic bacteria, yeasts and
moulds. In some case, essential oil like thymol and carvacrol were added to a bio-based
coating as antimicrobial agents and not as flavouring compounds 49,50. Ponce et al. 51 used
12/04/2014 12:05:39.
oleoresins containing both volatile and non-volatile compounds extracted from oregano,
rosemary, olive, capsicum, garlic, onion, and cranberries in edible films based on sodium
caseinate and/or carboxymethylcellulose and chitosan. These authors displayed that the use
of chitosan enriched with rosemary and olive did not introduce deleterious effects on the
sensorial acceptability of squash. Chitosan enriched with rosemary and olive improved the
antioxidant protection of the minimally processed squash, offering a great advantage in the
prevention of browning reactions which typically result in quality loss in fruits and
vegetables. These coatings provided both antioxidant and antimicrobial properties of
coatings at 1% oleoresin content without too much sensory disturbance.
Gelatinechitosan-based edible films incorporated with clove essential oil were elaborated
and their antimicrobial activity tested against six selected microorganisms: Pseudomonas
fluorescens, Shewanella putrefaciens, Photobacterium phosphoreum, Listeria innocua,
Escherichia coli and Lactobacillus acidophilus. The clove-containing film inhibited all
these microorganisms irrespectively of the film matrix or type of microorganism 52. The
effect on the microorganisms during this period was in accordance with biochemical
indexes of quality, indicating the viability of these films for fish preservation.
These authors also tested on 18 bacterial strains other essential oils: fennel (Foeniculum
vulgare Miller), cypress (Cupressus sempervirens L.), lavender (Lavandula angustifolia),
thyme (Thymus vulgaris L.), herb-of-the-cross (Verbena officinalis L.), pine (Pinus
sylvestris) and rosemary (Rosmarinus officinalis). Antioxidant properties as well as light
barrier properties of gelatin-based edible films containing oregano or rosemary aqueous
extracts have been assessed by Gomez-Estaca et al. 53. The essential oil polyphenols
protein interaction was found to be more extensive when tuna-skin gelatine was employed.
However, this did not clearly affect the antioxidant properties of the films, although it
could affect diffusion of phenolic compounds in the essential oil from film to food. The
light barrier properties were improved by the addition of oregano or rosemary extracts,
irrespective of the type of gelatine employed. The shelf-life of cold-smoked sardine
View Online
(Sardina pilchardus) was improved also by gelatine-based films used; induced reactions
and/or oxidation, to increase flavours shelf-life and to allow a controlled release.
Incorporation of small amounts of flavours into foods can greatly influence finished
product quality, cost, and consumer satisfaction
Carrageenans, alginates, whey proteins, soy proteins, starches, chitosans were tested for
coating of polymer films for providing antimicrobial and antioxidant properties,
encapsulating either terpenes, alcohols, or polyphenols as well as drugs for medical
purpose applications 18, 29,54,55. From Marcuzzo et al. 55, wheat gluten films could represent
an interesting opportunity as active packaging: they could retain and release aroma
compounds gradually, showing mechanical and nutritional properties different from those
of typical lipid ingredients usually used for terpene encapsulation. Hambleton et al. 54
showed that carrageenans were able to prevent flavour oxidation and retention during food
processing when retained in edible films/coatings.
the active compound release is activated by the moisture level of the packaged food as
soon as it was conditioned. Moreover, addition of antimicrobial-chitosan coating was able
to reduce gas permeability up to 1000 times.
Paper is a biodegradable material with versatile applications as bag, wrapper, copier paper
and photo print. Paper can also be utilized for agricultural and food purposes. Generally,
paper is coated with non-degradable polymer coating materials such as polyethylene, wax,
polyethylene terephthalate and polybutylene terephthalate to provide barrier properties 58.
These materials are difficult to dispose of and cause environmental problems. Thus,
introducing hydrocolloids (proteins) as a coating material on paper and paperboard has
been investigated recently. For example, corn zein protein coated on paper provided grease
resistance and was found suitable for wrapping sandwiches or fast food 59. An effort to
improve the paper properties by coating whey protein isolate (WPI) on paper also showed
grease resistance and enhanced the gloss without altering mechanical properties 60.
Paperboard coated with undenatured and denatured WPI had reduced oxygen permeability.
Furthermore, various hydrocolloid materials such as chitosan, WPI, whey protein
concentrate and wheat gluten, starches, cellulose derivatives, calcium caseinate, were
coated on paper. Coated papers have proven also their barrier efficiency against flavour
loss or off flavour penetration into foods 61,62
High barrier sustainable composite films and materials (multilayers, charged polymers,
fibres etc.)
View Online
The appropriate packing atmosphere in food packaging is needed to avoid colour or taste
deviation, oxidation of fats, formation of microorganisms, or degrading nutrients. On the
one hand, to achieve these requirements coextruded or laminated multilayer plastic films
are widely used in the packaging industry whereby ethylene vinyl alcohol copolymers
(EVOH) are often used to obtain a sufficient oxygen barrier (Figure 3). On the other hand,
the plastic recycling processes generally consist of the preliminary separation of the
different types of polymers, the shredding of the plastic items, the washing of the resulting
flakes, and their recompounding and processing into new, lower demanding applications 28.
In such a context, while the combination of various layers is required for good food
preservation, the recyclability of multilayer packaging is compromised, as monomaterials
of high purity are needed for reprocessing and as the separation of synthetic polymer
constituents in coextruded or laminated films are often impossible. On the contrary,
separation of hydrocolloids from synthetic polymers is very easy using aqueous hot
solutions in which hydrocolloids are soluble and then can easily be separated from plastic
polymers. The hydrocolloid coating would represent a new application for the agrofood
industry waste while safeguarding the performance and enhancing the recyclability of
multilayer films.
Figure 2, shows multilayer composite films associating synthetic polymers and biopolymer
layers. Some proteins and polysaccharides displayed oxygen and carbon dioxide barrier
properties as low as the best synthetic plastics as EVOH or PVDC. But as for EVOH or
PVDC, they are highly sensitive to moisture. That is the reason why hydrocolloid layers
have to be sandwiched between two hydrophobic barriers having very low water vapour
permeability such as PET or PE. So recently, Bugnicourt et al. 28 and Kurek 18 prepared
12/04/2014 12:05:39.
100000
Oxygen permeance at 25C and 50% RH (cm 3 /(m.d.bar)
10000 Paper
LDPE
HDPE Cellulose
1000 PP PC
acetate EVA
PA12 Wax paper
PA11 Cellophane
100
Soy
10 PET PA6 Protein
PE/Chitosan/PE Chitosan
PET/Whey
1 protein/PE
PP/PVDC/PE PE/EVOH/PE
0,1
PET/chitosan/PE
0,01
0,01 0,1 1 10 100 1000 10000
Water vapour transmission rate at 25C - 0-85%RH (g/(m.d))
Hydrocolloids are promising materials for future packaging systems, providing both barrier
and active properties and in a sustainable way. They are also very recently considered for
the development of sensors for monitoring food quality and safety as they could be used as
support polymers. More and more sensors are being developed and studied to
communicate real-time information on the freshness or safety of packaged food to the
consumers. Indeed, because they are edible, hydrocolloids could contribute to apply
moisture indicators, ripeness sensors, contamination sensors, oxidation or oxygen
consumption indicators, cold chain rupture indicator, etc inside the package and then in
potential contact with foods. So, the future of hydrocolloids for packaging application is of
great potential and realistic.
References:
1 Debeaufort, F., Quezada-Gallo, J.A., Voilley, A.(1998) Edible films and coatings :
"tomorrow's packagings ". Critical Review in Food Science and Nutrition, 38(4), 299-
313
2 Desobry, S., Debeaufort, F. (2011) Encapsulation of Flavors, Nutraceuticals and
Antibacterials. in Edible Films and Coatings for Food and Other Applications. J Bai, R
Hagenmaier and L Baldwin Eds, Technomic Publishing Co., London, pp 334-366.
3 Okuyama, K., Noguchi, K., Miyazawa, T., Yui, T., & Ogawa, K. (1997). Molecular and
crystal structure of hydrated chitosan. Macromolecules, 30, 58495855
12/04/2014 12:05:39.
4 Kurek, M., Guinault, A., Voilley, A., Gali, K. and Debeaufort, F. Effect of relative
humidity on carvacrol release and permeation properties of chitosan based films and
coatings (2012). Food Chemistry in press,DOI j.foodchem.2012.11.132
5 Williams, P. A.; Phillips, G. O. (2000) In Handbook of hydrocolloids; Phillips G.;
Williams P.; Eds.; CRC Press, Cambridge, England, 2000; pp 1-19
6 Mariniello L., Giosafatto C.L., Di Pierro P., Sorrentino Angela, Porta R. (2008) Use of
the enzyme transglutaminase to prepare hydrocolloid-based edible films suitable for the
food industry. J. Biotechnol., 136, 329-336
7 Choudhary D. R., Patel V., Patel H., Kundawala A. J. (2011). Exploration of Film
Forming Properties of Film Formers Used in The Formulation of Rapid Dissolving
Films. Int. J. Chem. Tech. Res., 3(2), 531-533.
8 Vicente A. A. (2011). Can natural hydrocolloids be used for nanotechnological
application in foods. PhD Thesis, University of Minho, Portugal.
9 Bayramoglu B., Stroeve P. (2010). Ultrathin edible coatings by the layer-by-layer (LbL)
assembly technique. Oral presentation at Institute of Food Technologists Annual
Meeting, July 18th,
10 Krochta, J. M. (2002) Proteins as Raw Materials for Films and Coatings: Definitions,
Current Status, and Opportunities In Protein-based Films and Coatings. Gennadios A.;
Ed.; Boca Raton, CRC Press, pp 1-41.
11 Skurtys O., Acevedo C., Pedreschi F., Enrione J., Osorio F., Aguilera J. M. (2010) Food
Hydrocolloid Edible Films and Coatings in Food Hydrocolloids: Characteristics,
Properties and Structures, Hollingworth C.S. (Ed.), London, Nova Science Publishers,
pp. 41-80.
View Online
12 Lacroix M., Cooksey K. (2005). Edible films and coatings from animal origin proteins.
In Innovations in Food Packaging, J.H. Han (Ed.), Elsevier, New York, pp301-317
13 Wu Y , Weller CL , Hamouz F , Cuppett SL and Schnepf M (2002) Development and
application of multicomponent edible coatings and films: A review . Adv. Food Nutr.
Res , 44 : 347 394 .
14 Letendre M , DAprano G , Lacroix M , Salmieri S , St-Gelais D (2002).
Physicochemical proper-ties and bacterial resistance of biodegradable milk protein films
containing agar and pectin . J.
Agric. Food Chem. 50, 6017 6022.
15 Ressouany M , Vachon C , Lacroix M (1998) . Irradiation dose and calcium effect on
the mechanical properties of cross-linked caseinate films . J. Agric. Food Chem., 46,
1618 1623.
16 Thakur BR , Singh RK , Handa AK (1997) . Chemistry and uses of pectin- a review .
Crit. Rev. Food Sci. Nutr., 37, 47 73.
17 Peroval C., Debeaufort F., Despre D., Voilley A. ( 2002). Arabinoxylan-lipid-based
edible films : 1 Effect of lipid type on water vapor permeability, film structure, and
others physical properties. J. Agric. Food Chem., 50(14), 3977-3983.
18 Fabra MJ, Chambin O., Assifaoui A., Debeaufort F. (2012). Influence of temperature
and salt concentration on the release in liquid media of aroma compounds encapsulated
in edible films. J. Food Eng., 108, 3036
19 Kokoszka S., Debeaufort F., Lenart A., Voilley A. (2010) Liquid and vapour water
transfer at interface related to structure properties of whey protein isolate/rapeseed oil-
based emulsion based films. J. Sci. Food Agric. 90, 16731680.
12/04/2014 12:05:39.
20 Galus S. Mathieu H., Lenart A., Debeaufort F. (2012). Effect of modified starch or
maltodextrines incorporation on the barrier and mechanical properties, moisture
sensitivity and appearance of soy protein isolate-based edible films. Innov. Food Sci. &
Emerging Technol., 16, 148-154
21 Garca M.A., Pinotti A., Martino N., Zaritzki N.E. (1999). Characterization of
composite hydrocolloid films. Carbohydr. Polym., 56, 339-345.
22 Banker GS (1966) Film coating theory and practice . J. Pharm. Sci., 55, 81 89.
23 Krochta JM , De Mulder-Johnston C (1997) . Edible and biodegradable polymer films.
Challenges
and opportunities . Food Technol. 51 , 61 74.
24 Debeaufort F., Voilley A. (2009) Structure and function of wax, fat and lipid-based
edible films and coatings. In Edible Films and Coatings for Food Applications, M.E.
Embuscado, K.C. Huber (Eds.), Publisher Springer Science, pp 139-172
25 Debeaufort F., Martin-Polo M., Voilley A. (1993). Polarity, homogeneity and structure
affect water vapor permeability of model edible films. J. Food Sci., 58(2), 426-434.
26 Krochta JM (1997) . Edible protein films and coatings. In Food proteins and their
applications in foods, S Damodaran , A Paraf (Eds.),. Marcel Dekker , New York, NY,
pp. 529 549
27 Kampf and Nussinovitch (2000). Hydrocolloid coating of cheeses. Food Hydrocoll.,
14(6), 531-537
28 Bugnicourt E., Schmid M., McNerney O., Wildner J., Smykala L., Lazzeri A., Cinelli P.
(2013) Processing and Validation of Whey-Protein-Coated Films and Laminates at
Semi-Industrial Scale as Novel Recyclable Food Packaging Materials with Excellent
Barrier Properties. Adv. Materials Sci. Eng., 2013(1), 1-10.
29 Kurek M. (2012) Comprehensive study of the effects of formulation and processing
parameters on structural and functional properties of active bio functional properties of
View Online
37 Kim D.K., Lim J., Bae I.Y., Lee H.G., Lee S. (2011) Effect of hydrocolloid coatings on
the heat transfer and oil uptake during frying of potato strips. J Food Eng., 102, 317
320
38 Alimi B. A., Shittu T. A., Sanni L. O., Arowolo T. A. (2013). Effect of some
hydrocolloids as adjuncts on the quality of whole egg or egg white coated fried yam
chips. J. Food Agric. Environ., 11(1), 19-24.
39 Sanchez-Gonzalez L. (2010). Caracterizacion y aplicacion de recubrimientos
antimicrobianos a base de polisacaridos y aceites esenciales. PhD Dissertation,
Universidad Politecnica de Valencia, pp 1-309.
40 Burt, S., (2004). Essential oils: their antibacterial properties and potential applications
in foods-a review. Int. J. Food Microb. 94, 223253.
41 Rupasinghe, H.P., Boulter-Bitzer, J., Ahn,T., Odumeru, J., (2006). Vanillin inhibits
pathogenic and spoilage microorganisms in vitro and aerobic microbial growth in fresh-
cut apples. Food Res. Int., 39, 575580.
42 Friedman, M., Henika, P.R., Levin, C.E., Mandrell, R.E. (2004). Antibacterial activities
of plant essential oils and their components against Escherichia coli O157:H7 and
Salmonella enterica in apple juice. J. Agric. Food Chem., 52, 60426048.
43 Raybaudi-Massilia, R., Mosqueda-Melgar, J., Martin-Belloso, O. (2006). Antimicrobial
activity of essential oils on Salmonella Enteritidis, Escherichia coli, and Listeria
innocua in fruit juices. J. Food Prot., 69, 15791586.
44 Fenaroli, G. (Ed.). 1995. Fenarolis Handbook of Flavor Ingredients. CRC Press,
Florida.
45 Sanchez-Gonzalez L., Vargas M. , Gonzalez-Martinez C., Chiralt A., Chafer M.,
(2009). Characterization of edible films based on hydroxypropylmethylcellulose and tea
tree essential oil. Food Hydrocoll., 23, 21022109
View Online
46 Sanchez-Gonzalez L., Gonzalez-Martinez C., Chiralt A., Chafer M., (2010). Physical
and antimicrobial properties of chitosantea tree essential oil composite films. J. Food
Eng., 98, 443452
47 Rojas-Grau, M.A., Avena-Bustillos, R., Friedman, M., Henika, P., Martn-Belloso, O.,
McHugh, T., (2006). Mechanical, barrier and antimicrobial properties of apple puree
edible films containing plant essential oils. J. Agric. Food Chem., 54, 92629267.
48 Rojas-Grau, M., Raybaudi-Massilia, R.M., Soliva-Fortuny, R.S., Avena-Bustillos, R.J.,
McHugh, T.H., Martn-Belloso, O. (2007). Apple puree-alginate edible coating as
carrier of antimicrobial agents to prolong shelf-life of fresh-cut apples. Postharvest Biol.
Technol., 45, 254264.
49 Ben Arfa , A., Chrakabandhu, Y., Preziosi-Belloy, L., Chalier, P., Gontard, N. (2007).
Coating papers with soy protein isolates as inclusion matrix of carvacrol. Food Res. Int..
40, 2232.
50 Del Nobile, M.A., Conte, A., Incoronato, A.L., Panza, O. (2008). Antimicrobial
efficacy and release kinetics of thymol from zein films. J. Food Eng., 89, 57 63.
51 Ponce A G., Roura, S.I., del Valle, C.E., Moreira, M.R., (2008). Antimicrobial and
antioxidant activities of edible coatings enriched with natural plant extracts: In vitro and
in vivo studies Postharvest Biol. Technol., 49, 294300.
52 Gmez-Estaca J., Montero P., Fernndez-Martn F., Alemn A., Gmez-Guilln M.C.
(2009). Physical and chemical properties of tuna-skin and bovine-hide gelatin films
with added aqueous oregano and rosemary extracts. Food Hydrocoll., 23(5), 1334-1341.
53 Hambleton A., Debeaufort F., Beney L., Karbowiak T., Voilley A. (2008). Protection of
active aroma compound against moisture and oxygen by encapsulation in biopolymeric
12/04/2014 12:05:39.
HEALTH ASPECTS
12/04/2014 12:05:41.
View Online
View Online
1.INTRODUCTION
Modern diets are considered to contribute significantly to the increase in obesity, diabetes and
cardiovascular diseases across a wide range of social strata.1 Recently, a great deal of focus
has been given to the consumption of processed foods as they often contain high levels of salt,
fat and sugar. Long-term consumption of food high in fat or sugar, are mainly responsible for
excess energy intake, having great impact on bodyweight. Overconsumption of salt is
associated with high blood pressure (hypertension), a major risk factor for cardiovascular
disease. 1 In addition processed foods are digested too rapidly and as consequence they exert a
12/04/2014 12:05:41.
3. IN-BODY FUNCTIONALITY
3.1 Colloidal bases of oral processing and sensory perception
12/04/2014 12:05:41.
It might appear obvious to the reader that in order to promote health, food must be eaten. This
implies that as a prerequisite every food material needs to be palatable otherwise no one will
eat it. The starting point is therefore to consider what happens to foods once they enter the
mouth with the emphasis on the mechanism underlying food perception. The perception of
food in the mouth arises from the physiological processes together with the dynamic changes
in food structures.12 Food is broken down through mastication, mixed with saliva and
equilibrated to the physiological temperature of the oral cavity.13 Significant efforts have been
In-body
Sensory
Texture
Ingredients Flavour
Mouth
Food Food
Structure Functionality
Process Gut
Self-assembling
bio-availability
Figure 1: Schematic representation of the design microstructure approach for health and
wellbeing.
View Online
devoted towards understanding the relationship between rheological properties and sensory
perception. In the past, shear flow was assumed to be most relevant in the mouth and shear
viscosity at either 50 s-1 or 10 s-1 was found to correlate with thickness perception.14, 15
However, this approach has been criticised as the flow may be turbulent as well as be time-
dependent.16 In addition, as the food is broken down and mixed with saliva a thin-film of food
is formed between the tongue and palate, where friction and lubrication of oral surfaces
become progressively more important than bulk rheological properties.17 Lubricating
properties are particularly relevant for reduced fat products as the textural perception of fat
related attributes, such as creaminess and fatty mouth-feel, are inversely related to oral
friction, as initially proposed by Kokini.18 Remarkably, only a small amount of oil is needed to
significantly reduce the friction coefficient.19 Work by Malone et al.,20 studied the tribological
behaviour and sensory perception of oil-in-water emulsions of similar viscosity (100 s-1), but
varying in fat content (Figure 2a). The friction coefficient of oil-in-water emulsions with a fat
content above 15% shows little, if any, difference to the friction coefficient of pure sunflower
oil. The lubrication mechanism of oil-in-water emulsions at low speeds is controlled by the
formation of a continuum oil film in the contact zone between the two surfaces due to
droplet coalescence. At high speeds the rheology of the emulsions dominates the frictional
response as water enters in the gap. So, knowing the lubricating mechanism enables design of
oil in water emulsions with a controlled stability towards coalescence to promote the
formation of the continuum oil film which lowers the friction between oral surfaces to provide
enhanced fat perception. To this end one of the important factors is the type of emulsifier used.
Figure 2b shows the effect of the different emulsifiers on the rate of coalescence of oil-in-
water emulsions. The rate and extent of coalescence observed for the system emulsified with
gelatin is much lower than when monopalmitin is used. This occurs due to the thick interfacial
12/04/2014 12:05:41.
200
oil
0.091 180
55% no surfactant
30% 160
Friction coefficient ()
0.076 15%
140
Droplet size (d 3,2)
1% 1% monopalmitin
0.061 0% 120
100
0.046 1% gelatin
80
0.031 60
40
0.016
20
0.001 0
1 10 100 1000 0 4000 8000 12000
Figure 2: (a) Friction coefficient versus entrainment speed for a series of iso-viscous oil-in-
water emulsions (from Malone et al.,20. (b) Droplet coalescence (inflow) for emulsions
containing different emulsifiers (from Norton et al.,21).
View Online
Yet an important aspect to understand emulsion lubrication in the mouth is the influence of
saliva. Dresselhuis et al.,22 reported an increase in friction coefficient of an oil-in-water
emulsion stabilised by whey protein in presence of human saliva. This effect depends on the
adherence of proteins onto the two surfaces and interaction with saliva (glycoprotein). Further
support to this argument comes from the work by Vardhanabhuti et al.,23 where a decrease in
the lubricating properties of saliva upon the addition of whey protein was ascribed to the
interactions between positively charged -lactoglobulin (pH 3.5) and salivary proteins. The
generic conclusion that emerges from the above is that the knowledge on the mechanism
underlying sensory perception and how it is controlled by food microstructure is required to
design healthier food where palatability and acceptance are critical considerations.
intestine.27 This decreases fat absorption rates and so reduces the fat intake from foods.28 Ca
and Mg ions are associated with bitter, metallic or off tasting flavours when added to foods so
these will need to be delivered to the stomach or small intestine through some form of
colloidal structuring. This offers a possible technique by which reduction of fat absorption
efficiency can be controlled in the digestive tract. The microstructures of food and emulsions
have been shown to impact both digestion and availability of nutrients for absorption.
Emulsion droplet size and the emulsifiers used in stabilisation has been shown to alter milk fat
digestion.29 Reviews on encapsulation30 and emulsification25 of food and their impacts on
digestion have been produced by others. These variations could be used to develop healthier
foods with minimally altered ingredients. A detailed understanding of the physiology of the GI
tract allows developments in the design of food microstructures to provide health promoting
properties to foods. Inclusion of health promoting agents, self structuring for satiety and
limitation of nutrient absorption can all be developed from this knowledge.
4. REDUCTION OF FAT
4.1 Emulsions
Emulsions have been reported to behave sensorially and tribologically like pure oil if there is
no release of water and they contain more than 20% oil.31 This incorporation of an additional
phase allows the reduction of total fat in the system due to use as a filler. Figure 3 shows
confocal micrographs of a double emulsion in which the internal phase (figure 3b) shows
View Online
Figure 3: (a) Water in Oil (W/O) emulsion (b) Oil in Water in Oil (O1/W/O2) double emulsion
(Images provided by Lucie Villedieu and Robin Hancoks).
View Online
comparison to the time scale of the conformational ordering, then thermodynamically driven
spherical particles, of ~ 1 m in size, can be formed.42 Whilst rapid ordering favours particle
growth they quickly reach dimensions that are increasingly affected by the applied shear
forces. In this case the equilibrium particle size is determined by the mechanical break-up and
large (>100 m) anisotropic particles are formed.43 A wide range of physical behaviours can
emerge from varying the intrinsic properties of the gel particles such as size and rigidity. A
detailed description of the rheological and tribological properties of -carrageenan fluid gels
have been recently reported by Garrec et al.42 Figure 4a shows the effect of concentration on
the flow curves of N-carrageenan sheared gels. Both samples exhibit a shear-thinning type of
behaviour and the viscosity increases with increasing polymer concentration which is ascribed
to both increased particle elasticity and volume fraction. The existence of an apparent yield
stress is attributed to the formation of bridges between adjacent particles mediated by the
ordering of remaining disordered chains (at the particle surface) after processing. The
temperature at the end of sheared gel production controls this behaviour with high
temperatures (close to the disorder-order transition temperature) promoting re-ordering as
shear ceases and as consequence the yield stress increases. Figure 4b shows the effect of -
carrageenan sheared gels on the friction coefficient between soft-(PDMS) surfaces. The low
contact pressures of PDMS are thought to be representative of tongue on palate interactions
during food consumption. The presence of the -carrageenan gel particles reduces the friction
coefficient as the particle rigidity increases. This is because rigid particles reduce surface-
surface contact due to their resistance to compressive forces thus providing lubrication.
Lubricating behaviour is also affected by the ratio of particle diameter to surface roughness.
When -carrageenan particle size is equivalent to the surface roughness dimensions the
particles fit in the asperities of the surface thus reducing friction even at low speeds (Figure
12/04/2014 12:05:41.
4b).
Figure 4: Effect of concentration used to prepare -carrageenan fluid gel on (a) flow curves
and (b) lubricating properties [adapted from Garrec et al.,42, 44].
View Online
A number of studies on the effect of particles on fat-related sensation attributes using calcium
carbonate, fused alumina, silica, dioxide and polymer-systems have shown that large, hard,
sharp particles produce a more rough, gritty and unpleasant sensation than small, soft and
smooth particles.45, 46 With these developments it appears that production of very small
spherical particles small enough to not be sensed in the mouth with the required rigidity to
induce surface separation would allow the effective use of sheared gel particles as fat replacers
where a reduction in the friction coefficient is a critical consideration.
sample flow history can be used to kinetically trap flow induced structures, the morphology of
which is determined by the thermal history and process shear conditions.50 Bulk rheological
properties depend on the rheology of the continuous phase, volume fraction ratio, dispersed
phase shape as well as de-swelling from solvent partition between the two components.51
Shear-induced phase inversion has also been reported52 and it may have important
implications as it occurs at shear rates applied in the oral cavity. Figure 5b shows viscosity
evolution at different shear rates for a series of LBG/NaCas mixtures. The peak viscosity
observed at high LBG volume fraction is due to the larger size of the caseinate inclusions due
to shear induced coalescence.
View Online
Figure 5: (a) Phase diagrams of LBG/NaCas (LBG shows as black in the micrograph). (b)
Flow curves for blends of LBG/NaCas as the volume fraction of the polysaccharide phase is
progressively increased from 0 wt% (pure protein) to 75 wt%. (adapted from Spyropoulos
et al.,52).
Continued coalescence leads to the completion of the phase inversion process, this does not
occur when LBG is the dispersed phase. This is because phase inversion occurs to lower the
12/04/2014 12:05:41.
energy of the system, while under shear, where the phase with the highest viscosity (LBG)
becomes the dispersed phase. An understanding of the design principles of biopolymer
mixtures has allowed the development of low fat spreads which acceptably mimic the
spreading and mouth feel properties of a full fat product with a microstructure made of
gelatine particles in a matrix of gelled maltodextrin.53 Here the crystalline nature of
maltodextrin networks is crucial to mimic the mechanical properties of the fat crystal network.
Other examples of the use of biopolymer mixtures to form structure to control texture of low
fat products can be found in many other products such as ice creams.54 Limitations in the use
of biopolymer mixtures to control functionality arise from our lack of knowledge on the
molecular dynamics of those systems
5. SALT REDUCTION
Salt plays many roles in food not simply as a flavouring agent, modification of flavour,
enhancement of functionality and physical properties, prevention of fermentation and
enhancement of shelf life, reduction of salt in food impacts on all of these aspects. 55, 56 The
replacement of NaCl with potassium or magnesium salts in bread has been reported to result in
unpalatable metallic, bitter and off flavours when above 20% of the total salt is replaced.57 The
size of salt crystals has been found to influence the rate of salt delivery as a flavour.58, 59
Smaller crystals dissolve more rapidly with increased surface area to volume ratio, resulting in
a higher sodium concentration in the saliva. Larger crystals while dissolving more slowly
prolong the duration of the flavour release. Prolonged release requires less salt to be used for a
View Online
similar flavour, due to higher sensitivity to the initial lower salt level; following rapid release
higher concentrations are required to taste equally as salty, rapid release provides an
immediate indicator of the food flavour. Patented size distributions have been developed to
mask undesirable flavours.60 A bimodal distribution of NaCl to provide both a rapid and
sustained flavour release, with the addition of another bitter salt with an intermediate size
reduced the total quantity of NaCl required for an equivalent saltiness of food.60 Food
structure, composition salivation and mastication all influence the rate of salt release from a
bolus in addition to crystal size. The fraction of salt tasted has been reported to be low in some
foods.61 Control of saltiness while reducing sodium levels by altering salt crystal size is only
applicable to dry foodstuffs where the crystals remain intact prior to consumption. Pulses of
flavour stimuli like salt allow the perception upon consumption to be increased for the amount
of salt consumed; a continuously present taste reduces saltiness perception over time. 55
Pulsing therefore allows the use of less salt to create the same flavour profile, this idea has
been developed into inhomogeneous distributions of salt in foodstuffs which provide an
increased perceived saltiness while maintaining or reducing total salt levels.62-64 A number of
patents have been issued in reduced salt products.59, 60, 65-67 Liquid or moist foodstuffs typically
contain salt evenly distributed throughout the food due to dissolution in the moisture present
preventing pulsed delivery. Work to overcome this has been undertaken with fillers, emulsions
and encapsulation to create inhomogeneous distributions of salt in food microstructure.
Particulates in instant soup have been reported as producing this inhomogeneity.64
Encapsulation of salt in fat allowed creation of inhomogeneous distributions in bread, even
though bread dough is a moist system.63 Encapsulation of seasonings has been patented but is
a wide area of continuing study.67 Trapping of active ingredients in the interior of double
emulsions allows protection or triggered release of the material contained in the interior phase,
12/04/2014 12:05:41.
this could be applied to salt. Localisation of the salt allows inhomogeneous distribution of salt
within a product if the emulsion is stable, extending inhomogeneous distributions to fluid
systems. This has been reported by Frash-Melnik et al.35, 36 The creation of an encapsulating
shell ensures that salt is inhomogenously distributed until breakage of the encapsulation.
Pickering stabilisation in foods has been the subject of recent review 68 as an emulsification
methodology which exhibits promise for control of encapsulation and microstructure. A
number of methods to create inhomogeneous distributions of salt both spatially and temporally
have been discussed here. Encapsulation for liquid or moist systems and trapping to control
the release of salts exhibit promise in the development of reduced salt foods through
microstructural design.
6. SUGAR
Sugars have two major functions in foods as a flavour (sweet) and to influence the structure
and moisture of foodstuffs.69 Replacement of flavour sugar has been undertaken with a range
natural and artificial compounds which will not be considered in this review.70-73 The use of
microstructures to reduce sugar levels with increased perception and reduction of sugar
required for structuring are being undertaken with colloidal structures in foods.74 Like salt
which has already been discussed, temporal and spatial inhomogeneity has been found to
increase perception of sweetness from sugar.73 With sugar able to be trapped in hydrocolloids
like gelatin, structures have been made to enhance sweetness with reduced overall sugar
content.64, 75 Little work has been reported on the reduction of structural sugar in foods. The
View Online
creation of microstructures with new materials has been proposed one of which is the use of
clays. These unusual materials require a radical rethink of how foods are constructed and what
will be acceptable to consumers.
7 SELF STRUCTURING
Novel food products with high satiating capacity are of significant interest. Prolonged
satiation will reduce the desire to eat further therefore improving the control over the appetite
and reducing daily food intake.76 Evidence has shown that an increased sense of satiety can be
created using colloidal structures. Viscous fluids thickened with hydrocolloids provide a
sensation of fullness which is associated with delayed gastric emptying.77 The underlying
hypothesis is that increased viscosity reduces flow out of the stomach and also that the
dilution/hydration of the hydrocolloids in the stomach increases the gastric volume to be
emptied.77-79 However, viscous beverages may be unpalatable and unpopular with consumers.
An alternative mechanism by which hydrocolloids can contribute to satiety is via gelation.
Gelation can be triggered by the physiological conditions found in the gastric lumen, i.e. low
pH and body temperature. With gelation the greater sensation of satiety is thought to be a
combination of viscosity increase and mechanical properties of the gel.80 Strong gels would be
less likely to break up by the grinding action of the stomach which is reported to apply forces
in the order of 0.65N.78 Hoad et al.,80 showed that the sense of fullness was significantly
greater for meals containing the polysaccharide alginate (high and low guluronic acid content)
or guar than for the control (without any hydrocolloid). In addition, Norton et al.,81 reported
that a gelled alginate meal remained in the stomach for significantly longer time than a non-
12/04/2014 12:05:41.
gelled locust bean gum meal and its overall emptying rate was slower. The rate of gelation is
likely to play a major role in determining the effect of intragastric gelation on satiety. The rate
of gelation should be fast enough to allow structuring while the digesta stay in the stomach.
However, if the gelation rate is too fast only part of the digesta will be trapped in the gel
structure, thus decreasing the impact on satiety. Norton et al.,82 showed that low-acyl gellan
forms strong elastic gels in acidic environments similar to those present in the stomach where
maximum gel strength was observed between pH 3 and 4. The author argues that the time
scale of low-acyl gellan gelation, lower than alginate, is suitable for controlling satiety.
Temperature-induced structuring could also be used to induce satiety. Recently, Sullo et al.,83
showed that a solution of new regioselective substituted cellulose ethers could be induced to
gel at a temperature of 37oC, changes to the alkyl moieties distributions on the cellulose chain
can control the gelation temperature for these systems. Hydrocolloid gelation in the stomach
shows potential for use as a method to reduce food consumption. This requires a detailed
understanding of the relationship between the fine chemical structure and the physico-
mechanical properties of the chosen hydrocolloids to generate the desired microstructures.
8.CONCLUSION
In this review we discussed recent advances in food colloids research which help to address
some of the health issues related to food consumption. The emphasis was placed on the design
of colloidal microstructures which impact on mouth-feel, perception, digestion, satiety and
nutrient bioavailability. We stressed the importance of understanding how food is digested and
View Online
taken up by the body as well as how colloidal systems affect transport and diffusion
phenomena relevant to in-body functionality. So, to what extent are these colloidal concepts
transferable to real food? To narrow the gap between fundamental works and direct
applications to real foods it is important scientists translate these scientific advances into
practical possibilities which require collaboration between different areas of food research,
food science, material science, engineering, food technology, chemistry, nutrition and
medicine.
References
1. A. d. l. Hunty, British Food Journal, 97 (9), 3 (1995).
2. I. Norton, S. Moore and P. Fryer, Obesity Reviews 8, 83-88 (2007).
3. S. Palzer, Trends in Food Science & Technology 20 (5), 194-200 (2009).
4. J. M. Aguilera, Journal of Food Engineering 67 (1-2), 3-11 (2005).
5. T. J. Foster, TECHNOFUNCTIONALITY OF HYDROCOLLOIDS AND THEIR
IMPACT ON FOOD STRUCTURE. (Royal Soc Chemistry, Cambridge, 2010).
6. A. M. Hermansson, M. Langton and C. Olsson, in Gums and Stabilizers for the Food
Industry 12, edited by P. A. Williams and G. O. Phillips (2004), pp. 3-12.
7. I. Norton, P. Fryer and S. Moore, AIChE Journal 52 (5), 1632-1640 (2006).
8. D. R. Jacobs and L. C. Tapsell, Nutrition Reviews 65 (10), 439-450 (2007).
9. P. W. J. Maijaars, R. J. R. van der Wal, E. A. Haddeman, H. P. F. Peters, C. Beindorff
and A. A. M. Masclee, European Journal of Gastroenterology & Hepatology 20 (7), A54-A54
(2008).
12/04/2014 12:05:41.
10. D. Kilcast and C. den Ridder, in Reducing Salt in Foods: Practical Strategies, edited
by D. Kilcast and F. Angus (2007), pp. 201-220.
11. J. Ubbink, Faraday Discussions 158, 9-35 (2012).
12. J. B. Hutchings and P. J. Lillford, Journal of Texture Studies 19 (2), 103-115 (1988).
13. J. Chen, Food Hydrocolloids 23 (1), 1-25 (2009).
14. F. W. Wood, in SCI Monograph (Society of Chemical Industry, London, 1968), Vol.
27, pp. 10.
15. F. Shama and P. Sherman, Journal of Texture Studies 4 (1), 111-118 (1973).
16. T. van Vliet, Food Quality and Preference 13 (4), 227-236 (2002).
17. J. Chen and J. R. Stokes, Trends in Food Science & Technology 25 (1), 4-12 (2012).
18. J. L. Kokini, Journal of Food Engineering 6 (1), 51-81 (1987).
19. R. A. de Wijk and J. F. Prinz, Food Quality and Preference 16 (2), 121-129 (2005).
20. M. E. Malone, I. A. M. Appelqvist and I. T. Norton, Food Hydrocolloids 17 (6), 763-
773 (2003).
21. I. T. Norton, F. Spyropoulos and P. W. Cox, Food Hydrocolloids 23 (6), 1521-1526
(2009).
22. D. Dresselhuis, H. J. Klok, M. C. Stuart, R. Vries, G. Aken and E. A. Hoog, Food
Biophysics 2 (4), 158-171 (2007).
23. B. Vardhanabhuti, P. W. Cox, I. T. Norton and E. A. Foegeding, Food Hydrocolloids
25 (6), 1499-1506 (2011).
24. E. Bauer, S. Jakob and R. Moseenthin, Asian-Australasian Journal of Animal Science
18 (2), 282-295 (2005).
View Online
1. INTRODUCTION
12/04/2014 12:05:43.
In the past decades, polysaccharides isolated from natural sources (herbs) have
attracted much attention due to their various types of bioactivities. Both publication and
citations in scientific journals related to bioactive polysaccharides have increased steadily
particularly since 2003 (Fig. 1).
Figure 1. Published items (a) and citations (b) in each year, with bioactive polysaccharide
as the topics in the database of web of science (http://apps.webofknowledge.com)
A variety of polysaccharides extracted from botanical plants and fungi have been
reported for their bioactivities. The claimed health benefits on humans include immune
View Online
regulation, anti-tumor, antioxidant activity etc. Polysaccharides from mushroom have been
reviewed by our group which attracted much attention 1. In the present paper, we mainly
focus on polysaccharides isolated from three sources: Dendrobium officinal, Cordyceps
sinensis and Ganoderma. (Fig. 2). The structural information, bioactive properties as well
as their structural & function relationships are discussed.
Figure 2: Photograph of Dendrobium officinal (a), Cordyceps sinensis (b) and Ganoderma
(c) (photographs b and c were taken from the internet: b:
http://baike.baidu.com/view/7002.htm?subLemmaId=7002&fromenter=Ganoderma+lucidum;
c: http://cordyceps-sinensis-mushroom.blogspot.com/).
View Online
2.3.Immune-stimulatory effects
It has been reported that polysaccharides from Dendrobium officinal could inhibit free
radicals including hydroxyl radicals, superoxide radicals, ABTS radicals etc. and
malondialdehyde (MDA). He et al. (2007) reported that one polysaccharide isolated from
the suspension-culture of Dendrobium officinal could significantly inhibit (P < 0.01)
hydroxyl radicals and superoxide anion radicals in vitro. It could also significantly inhibit
(P<0.01) the formation of MDA in mice liver homogenate caused by auto-oxidation,
FeSO4-induced oxidation, or H2O2-induced oxidation 9.
For in vivo tests, parameters that are usually considered, include the thymus index, T
lymphocyte transformation rate, tumor inhibitory rate, half value of hemolysin (HC 50),
activity of natural killer (NK) cells, spleen index, phagocytic rate of peritoneal
12/04/2014 12:05:43.
macrophages, serum IL-2, TNF-, IFN-, and MDA in animal models. It has been reported
that gavage-fed with Dendrobium officinal polysaccharides for tumor-bearing mice could
decrease tumor weight effectively 13. Previous research also suggested Dendrobium
officinal polysaccharides could improve T lymphocyte transformation rates, increase
natural killer (NK) cell activities, phagocytic rates of peritoneal macrophages and half
value of hemolysin (HC50) 13. In addition, Dendrobium officinal polysaccharides were also
reported to improve the levels of serum cytokines, including IL-2 and TNF- 14.
Bao et al., (2012) isolated two heteroglycans (PL-1 and PL-4) and one glucan (PL-3) from
the fruiting bodies of G. lucidum. Structural analysis indicated that PL-1had a backbone of
(1-4)--D-glcp residues and (1-6)--D-galp residues, while PL-4 comprised of (1-3)-,
(1-4)-, (1-6)--glcp residues and(1-6)--manp residues. The glucan was highly branched,
composed of (1-3)--D-glcp residues substituted at O-6 with (1-6)-glucosyl residues 25.
Dong et al (2012) reported one water soluble polysaccharide from the spores of G.
lucidum. The structural characterization demonstrated a -D-glucan featured by a
1,6-linked -D-Glcp backbone with different length of branches consisting of terminal and
1,4-linked Glcp residues, attached to the O-4 of alternative Glcp residues in the backbone
21
G. lucidum has been reported to possess many bioactivities such as immune regulation,
anti-tumor, anti-oxidant activity, hypoglycemic effect, antibacterial effect, reducing blood
cholesterol, liver-protecting effect, inhibiting angiogenesis, anti-fibrotic and anti-HIV
Li et al., (2007) reported that polysaccharides from the submerged culture broth of a
basidiomycete G. lucidum showed the ability to enhance the T and B lymphocyte
proliferation and exhibited lymphocyte activity in vitro. Animal trials using mice also
indicated that polysaccharide from G. lucidum significantly enhanced the T and B
lymphocyte proliferation and antibody production, increased the mass of spleen tissue; this
polysaccharide also showed a hepatoprotective activity in mice with liver injury 26.
Considerable scientific evidence has been collected regarding the anti-tumor effects of
polysaccharides from Ganoderma, both in vitro and in vivo. G. lucidum polysaccharide
could significantly reduce the tumor weight and volume in a dose-dependent manner 32.
Possible mechanisms have been proposed, among which two most likely mechanisms are
(1) the polysaccharide caused significant cytotoxicity in human tumor, which could be
achieved through apoptotic effects 34. (2) G. lucidum polysaccharides could activate the
host immune function, modify the redox system and enhance the immune response. 35.
View Online
G. lucidum polysaccharide was reported to increase the insulin levels and decrease
blood glucose in streptozotocin (STZ)-induced diabetic mice 39. G. lucidum polysaccharide
12/04/2014 12:05:43.
was also reported to be able to decrease total cholesterol (TC) and triglyceride in
high-fat-fed/streptozotocin (STZ)-diabetic rats 40.
Cordyceps sinensis (C. sinensis), known as winter worm summer grass in Chinese, is
mainly distributed in the prairie soil of above 2000 m on the Tibetan plateau in China.
Cordyceps were traditional treated as a herb and used as a lung protectorate and for
kidney improvement. Evidences of the beneficial health effects of C. sinensis have been
collected, which mainly includes immune modulating effects, anti-aging, anti-tumor,
anti-microbial and antioxidant activity 41-50. Other health benefits on humans such as
protection against renal toxicity, heart disease, liver disease, and respiratory disease were
also reported. 41. These beneficial effects have been partly attributed to the specific
structure of polysaccharides from C. sinensis.
Polysaccharides can be extracted from both wild C. sinensis and cultured C. sinensis.
Miyazaki et al. (1977) isolated a water soluble fraction (CS-I) from natural C. sinensis,
which was a highly branched galactomannan with mainly (1-2)- -linked-D-manp residues
as the backbone chain. The branches contained (1-3), (1-5), (1-6)-linked-D-galf and
(1-4)-linked-D-galp residues 42. Kiho et al. (1986) also isolated a galactomannan fraction
from C. sinensis with a weight average molecular weight of 23 kDa. It was composed of
View Online
D-mannose and D-galactose in the molar ratio of 3:5. Structural investigation indicated
that this polysaccharide contained a core of (1-2)- and (l-6)- linked -D-manp residues,
some of which were substituted at O-2 or -6 and at O-4 or O-6 with short chains of mostly
(l-5)-linked -D-galf residues and a small percentage of (l-6)-linked a-D-galp residues 43.
Due to environmental changes and over exploitation, C. sinensis from the wild is near
extinguished. As a result, mycelia fermentation of Cordyceps fungal is used as a substitute.
The structural features of polysaccharides from fermented C. sinensis can be classified into
three groups based on molecular weight and monosaccharide composition. The first group
has a relatively large molecular weight and contains mainly glucose, galactose and
mannose. For example, Kiho and his colleagues isolated a galactoglucomannan (CS-F10)
from cultured mycelium of C. sinensis which comprised of galactose, glucose and mannose
in a molar ratio of 43:33:24 with molecular weight about 15 kDa. Structural investigation
indicated that CS-F10 had a comb-type structure with -D-glcp residues at the terminal end
of the side chains. It also contained (1-5 and/or 6)-linked -D-galf residues, and
(1-2)-linked and branched -D-manp residues 44. Yan, Li, Wang, & Wu (2010) extracted a
polysaccharide (designated EPS-1A) with an average molecular weight of 40 kDa.
Structural analysis showed that EPS-1A was composed of glucose, mannose and galactose
at molar ratio of 15.2:3.6:1.0. The backbone contained (1-6)--D-glucose residues (77%)
and (1-6)--D-mannose residues (23%). Branches contained (1-6)--D-mannose residues
12/04/2014 12:05:43.
An acidic polysaccharide was also isolated from C. sinensis in mycelial culture. This
polysaccharide contained Glcp and GlcUp in a molar ratio of 8:1 with the average
molecular weight of 36 kDa. The backbone comprised of 3--D-Glcp residue, while two
branches, -D-Glcp and -D-GlcUp, were attached to the main chain by (16) glycosidic
bonds at every seventh -D-Glcp unit 46.
The third group contains mainly glucose as monosaccharide with a relatively small
molecular weight. An insoluble glucan-like polysaccharide (CS-Pp) was obtained from the
mycelia of the C. sinensis with the monosaccharide composition ratio of
Glc:Man:Gal=21:2:1. CS-Pp is a 1,3--D-glucan with some 1,6-branches 47. Nie et al.,
(2011) reported the structure of a bioactive hydrophilic polysaccharide fraction from C.
sinensis (CBHP): the backbone is composed of Glcp joined by 14 and 13 linkages,
with the branching points located at O-2 or O-6 of Glcp with - terminal-D-Glcpas side
chain 41. The proposed structure is shown in Figure 7.
2 2
4
G14G14G14G14G14G1(4G14G14G14G14G14G14G13G14G14G1)24G14G13G14G14G14G14G1
6 6 6 6
1 1 1 1
G G G G
Figure 8: CS inhibited UUO induced mice renal fibrosis. Control groups were sham
12/04/2014 12:05:43.
operated animals [Left] and UUO untreated animals [Middle]. In the treatment group
[Right] mice were treated with CS by oral gavage for three days prior to surgery (160
mg/kg, thrice per day). 7 days after UUO surgery, Massons trichrome-staining was used to
evaluate the extent of renal fibrosis which was subsequently quantified. This figure was
taken from Zhang et al, (2012) 49.
It was reported that crude polysaccharides from wild type C. sinensis and mycelia of C.
sinensis could induce macrophages from the mouse abdominal cavity to produce the tumor
necrosis factor (TNF-). According to Lin et al., (2004), polysaccharide from wild C.
sinensis was better than the mycelia of C. sinensis in the ability to induce macrophage from
the mouse abdominal cavity in vitro to produce TNF- 55. In addition, the acidic
polysaccharide (AEPS-1) from C. sinensis could significantly stimulated the release of
several major cytokines (TNF-, IL-1, IL-6 and IL-10), demonstrating an
immunomodulatory property 46.
6. CONCLUDING REMARKS
Figure 9. Triple helix structures of polysaccharide, derived from molecular modeling 65.
View Online
e.g. ethanol precipitation, ammonium sulphate precipitation, size exclusion and ion
exchange chromatography. However, it is always a challenge to conclude that no
interference from small molecules or other contaminants which could be associated with
the polysaccharides, either physically trapped or covalently linked. In addition, although
many in vitro and in vivo methods (models) are used to characterize the bioactivities of
polysaccharides, the accuracy and effectiveness for some of these methods are yet to be
validated. Each research group favors their individual method which makes it difficult to
compare the results from different labs. Therefore, the understanding of the relationships
between structure and bioactivity of these polysaccharides still has a long way to go.
References
1
Department of Food Science and Technology. 2Department of Biochemistry and
Biotechnology, Kwame Nkrumah University of Science and Technology, Kumasi.
Introduction
Cereal cell wall (13) (14) -D-glucans possess functionality in processed food systems
and impart physiological functionality from which health benefits are derived.1 The
conventional source of cereal -glucan is barley and oat which occur at 3-11 % and 3-7 %,
respectively.2,3 Other minor sources of glucan are wheat and rye 0.5 - 2 %2,4,5,6,7 and 0.12 to
5.4 % in sorghum.4,8,9 In a recent screening of tropical maize varieties, -glucan content of
three cultivars produced in Ghana ranged from 1.4 2.5 %.10 Functional characteristics of -
glucan include water binding ability and formation of viscous suspensions which exhibit
unique rheological properties, which are utilized in food formulations for texture11,12,13 and
can be exploited as fat replacers in baked goods.14
The rheological properties of oat and barley -glucan have been investigated in the
past.6,15,16,17,18,19 Barley and oat -glucan suspensions are non-Newtonian fluids which exhibit
a flow behavior characterized as pseudoplastic or shear thinning.19,20,21 There is however
dearth of information on rheological properties of maize -glucan. As a hydrocolloid, -
glucan, used as a fat replacer is expected to mimic the functionality of fat in baked products
through its viscosity enhancing effect, gel forming ability22,23 and water binding capacity. The
utilizability of maize -glucan as a fat replacer in pie crust has not been investigated. This
View Online
research seeks to investigate the rheological properties of aqueous dispersions of three maize
-glucan isolates and their functionality as a fat replacer in pie crust.
As yield stress, o was found to be zero from the plot of shear rate versus shear stress. In this
x
model V is the shear stress (Pa), J is the shear strain rate (s-1), K is the consistency
coefficient (Pa sn), and n is the flow behavior index (dimensionless). Logarithmic plots of
apparent viscosity () versus shear rate were used to determine the flow behavior and
consistency coefficient.26
Pastry preparation
Three sources of maize -glucan were evaluated as fat replacers. Reduced fat pie crusts were
formulated27 with maize -glucan (hereinafter referred to as MaiLean) as fat replacer at 15 and
20 % replacement. Ingredients and quantities used in each pastry variation are shown in Table
1. Full-fat pie crust served as control. Three samples of MaiLean originating from three maize
genotypes, Obatanpa GH (MaiLean OB), Abeleehi (MaiLean AB), and GH 9 (MaiLean
GH) were evaluated.
View Online
Sensory evaluation
Sensory attributes of reduced-fat pie crust were studied by means of a 33 factorial
12/04/2014 12:05:45.
experiment laid out in a completely randomized block design in three replications. Eighteen
semi-trained panelists were introduced to the control pie crust, as well as individual seven-
point hedonic rating scales for evaluating five attributes, namely, color, taste, crispiness,
brittleness, and overall acceptance. Color was assessed by visual inspection and taste was
assessed by the tongue. Brittleness was defined as the force required to break crust when
placed between the incisors. Eighteen panelist evaluated pie crust in a random order, rinsing
the mouth between evaluations. Perceived attributes were recorded on scorecards having the 7
point hedonic scale where 1 represented dislike extremely, 4 was neither like nor dislike and 7
represented like extremely.
Data analyses
A descriptive analyses encompassing means, median, standard errors and coefficient of
variation were computed. A one way analysis of variance (ANOVA) was carried out using
SPSS version 17, at a significance level of 5%. The Least Significant Dierence (LSD) test
was used to locate differences in means.
unstable, hence all viscosity measurements were performed at 10 % where stable viscosities
could be recorded. Figs. 1 and 2 show rheograms for -glucan dispersions. Viscosities of
maize -glucan was lower than that of oat dropping from 0.74 Pa s at the lowest shear rate of
6.4 s-1to ~0.006 Pa s at highest shear rate of 12,800 s-1, whereas oat -glucan viscosities
dropped from 3.6 Pa s to 0.77 Pa s at the same shear rates. The low viscosity of maize -
glucan compared to oat could be attributed to isolation method and/or the inherent molecular
structure of maize -glucan. Our inability to take viscosity measurements at concentration
below 5 % of oat demonstrates some degradation by the isolation methods. A linear
relationship has been established between viscosity, concentration and molecular weight of -
glucan isolated from oat and barley, as -glucan with degraded molecular structure exhibits
low viscosity than those of intact structures.28. Ahmad et al.29 recorded a viscosity range of
0.036 and 0.056 Pa s for less than 1% barley gum dispersion using a mild enzymic extraction
technique.
The rheological characteristics of maize -glucan were similar to that of oat (Fig. 1). Both
maize and oat gum dispersions exhibited three-stage viscous response when subjected to high
shear strain rate which is characterized by a Newtonian behavior at low shear rates between 0
and 6.4 s-1, a shear thinning range between 160 to 640 s-1, and finally, an infinite shear
viscosity at very high shear rate above 6,400 s-1 on account of the stages of conformational
rearrangement of the biopolymer that takes place during shearing.30 A similar observation was
made on barley gum dispersion31 that the rheological behavior of maize -glucan is identical
to that of oat and barley.
100000 1000
Shear stress (Pa)
10000 B
A
1000 100
100 10
10
1 1
1 100 10000 1 100 10000
Shear rate (s-1) Shear rate (s-1)
1000 1000
C
Shear stress (Pa)
D
Shear stress (Pa)
100 100
10
10
1
1 100 10000 1
Shear rate (s-1) 1 10 100 1000 10000 100000
Shear rate (s-1)
Figure 1: Rheograms of -glucan dispersions. (A) 5 % oat, (B) 10% MaiLean GH, (C) 10 %
MaiLean OB, (D) 10 % MaiLean AB.
View Online
10000 1000 B
A
1000
100
100
10
10
1 1
1 100 10000 1 100 10000
Shear rate (s-1) Shear rate (s-1)
1000 1000
Apparent viscosity (Pa s)
10 10
1 1
1 100 10000 1 100 10000
Shear rate (s-1) shear rate (s-1)
12/04/2014 12:05:45.
Figure 2: Apparent viscosity as a function of shear rate for -glucan dispersions. (A) 5 %
oat, (B) 10% MaiLean GH, (C) 10 %, MaiLean OB, (D) 10 % MaiLean AB
In each rheogram, the shape of the curve is typical of nonNewtonian fluid, characterized by
immediate deformation of the fluid under small stress, hence a yield stress of zero, followed
by a continuous change in the ratio of shear stress to shear strain rate and flow behavior index
of less than unity, conforming to the Power Law model.
A plot of apparent viscosity against shear rate (Fig. 2) demonstrated high apparent viscosities
which decreased rapidly with increase in shear strain rate, followed by a gradual drop over
time. Fitting the Power law model gave a flow behavior index, n of <1.0 and consistency
coefficient of >1.0 (Table 2), hence, maize -glucan dispersion can be described as shear-
thinning or pseudoplastic fluid. Studies by Ghotra,19 Doublier and Wood,20 and Lazaridou et
al.21 on the rheological behaviour of oat and barley -glucan revealed flow behaviour index
less than one thus attesting to pseudoplasticity of cereal -glucan. Marcotte et al.32 stated that
the Power Law model describes the pseudoplastic behavior of gums.
For maize -glucan, R2 values of the regression model of approximately 0.75 demonstrates
that shear strain rate explains about 75 % of the variation in shear stress, whereas for oat -
glucan, 81% of the variation in shear stress is explained by shear rate (Table 2). Many
polymer melts and solutions exhibit flow behaviour in the region 0.3 to 0.7 depending on the
concentration and molecular weight of the polymer.33,34 Doublier and Wood20 reported a flow
View Online
Table 2: Flow behavior (n) and consistency coefficient (K) at 5% oat and 10% maize gum
dispersions at shear rates of 0 to 12,800 s-1 at 27oC.
Beta-glucan dispersion at 10 %
Parameter MaiLean OB MaiLean AB MaiLean GH Oat
Power Law parameter Estimate
Yield stress 0 0 0 0
Flow behavior index (n) 0.50 0.21 0.480.24 0.520.22 0.910.58
Consistency coefficient (K) 1.150.55 1.230.54 1.170.57 2.131.00
Mean square 6.71 6.24 7.21 22.28
Standard error 0.01 0.01 0.17 0.04
R2 0.75 0.74 0.76 0.81
behaviour index 0.7 for oat and barley -glucan gums at concentrations above 0.5%. The flow
behavior index for 10 % maize -glucan showed 0.48 for Mailean AB,, 0.50 for Mailean
OB, and0.52 for Mailean GH while that of 5 % oat was 0.91 indicating the pseudoplastic
nature of the -glucan dispersions. From the forgone discussion, maize beta-glucan may also
be utilized as hydrocolloids in food formulations.
baked products.36 Of the total number of panelists who performed the sensory evaluation, 60%
were males and 40% were females. The colour of reduced-fat pie crust was similar to that of
full fat product. Table 3 shows the descriptive statistics for brittleness and acceptance of -
glucan fat-replaced and full fat pie crust. Increase in fat replacement above resulted in poorer
sensory attributes of pies; however, full fat and 15 % fat-replaced pie crust were comparable.
Table.3: Descriptive statistics for brittleness and acceptance of -glucan fat-replaced pie
crust.
Statistic Brittleness Overall acceptance
Full fat 15 % 20 % Full fat 15 % 20 %
Number analyzed 54 54 54 54 54 54
Mean score 5.30a 5.11a 4.78b 5.70a 5.60a 5.30b
Median 5 5 5 6 6 5
Standard 1.06 1.28 1.50 1.19 1.04 1.44
Deviation
SE 0.14 0.17 0.20 0.16 0.14 0.20
Variance 1.12 1.65 2.25 1.42 1.08 2.00
CV (%) 20.2 25.67 30.01 19.86 17.35 26.10
Minimum 3 2 1 1 2 2
Maximum 7 7 7 7 7 7
Numbers in rows followed by different superscripts are significantly different (P<0.05). CV = coefficient of
variation, SE= standard error of mean
View Online
A validation of the sensory data was performed at a confidence interval of 95%. For
brittleness, an unpaired t-test demonstrated no significance difference (P<0.05) between full
fat and 15 % fat-replaced pie crust. Fifty percent of panelist liked the brittleness of both full-
fat and 15 % fat-replaced pie crust equally, producing a mean scores close to 5 and a median
of 5. Similarly, regarding overall acceptance, there was no significant difference between full
fat and 15 % fat replaced pie crust (Table 2) In contrast, 20 % fat-replaced pie crust produced
poorer sensory scores which could be attributed to the insufficient amounts of fat needed to
inhibit the development of tough gluten networks.36 This observation is corroborated by
Chysirichote37 who stated that, fat has the ability to coat the surface of flour particles thus
disrupting the development of tough gluten proteins for softer and tender texture in baked
products.
Of the three maize -glucan isolates, MaiLean OB and GH at 15 % replacement in pie crust
gave sensory attributes similar to that of full-fat pie crust. However, MaiLean AB at the same
replacement level scored poorer in brittleness and acceptability. Functionality of hydrocolloids
as fat replacers is reflected in their water binding capacity. In another study (results not
shown) the water binding capacity of Mailean AB was significantly lower than the other
isolates.
Conclusion
The forgone study on rheological properties of maize -glucan revealed a pseudoplastic flow
behavior of dispersions, similar to the rheological properties of oat. The maize -glucan
isolates were found to possess characteristics that could be exploited as fat replacer in pie crust
formulation. Of the three isolates MaiLean OB and GH performance at 15 % fat replacement
gave pie crusts with similar sensory attributes which were not different from full-fat products
12/04/2014 12:05:45.
(P>0.05).
Table 4: Descriptive statistics for15% fat-replaced -glucan pie crust with MaiLean products.
Regarding brittleness and acceptance, no significant difference was observed between fat-
replaced pie crust developed from OB and MaiLean GH, whereas pie crust produced from
MaiLean AB showed significantly poorer scores at both 15 and 20 % replacement. . Please
assign reason to poor performance of MaiLean AB.
References
1. Miller, S.S. and Fulcher, R.G. 1994. Distribution of (13),(14)--D-glucan in
kernals of oats and barley using microspectrofluormetry. Cereal Chem. 71:64-68.
2. Bhatty, R. S. 1993. Extraction and enrichment of (13),(14)--D-glucan from
barley and oat brans. Cereal Chem. 70:73-77.
3. Bhatty, R. S. 1995. Laboratory and pilot plant extraction and purification of -glucans
from hull-less barley and oat bran. J. Cereal Sci. 22: 163170.
4. Palmer, G. H. and MacKenzie, C. I. 1986. Levels of alkali-soluble -D-glucans in
cereal grains. J. Inst. Brew. 92: 461-462.
5. Beresford, G. and Stone, B. A. 1983. (13),(1 4)--D-glucan content of Triticum
grains. J. Cereal Sci. 1:111-114.
6. Skendi, A., Biliaderis, C. G., Lazaridou, A. and Izydorczyk, M. S. 2003. Structure and
rheological properties of water soluble -glucans from oat cultivars of Avena sativa
and Avenabysantina. J. Cereal Sci. 38:15-31.
7. Cui, S. W. and Wang, Q. 2009. Structurefunction relationships of cereal -glucans.
In: Oral presentation at the 8th international hydrocolloids conference, held in
Trondheim, Norway
8. Ogbonna. A. C. and Egungwu, A. L. 1994. -glucan degradation in malting sorghum.
12/04/2014 12:05:45.
1 INTRODUCTION
Hydrocolloids are ingredients that food manufacturers use extensively to develop and
maintain a variety of functional properties in a wide range of food systems. Examples
include maintaining colloidal stability in beverages, inhibiting ice recrystallization in
frozen dairy products, improving texture in dough and increasing viscosity in emulsions.
The benefits of foods formulated with hydrocolloids are multifaceted as they have also
been linked to a range of physiological responses in humans, through their ability to act as
12/04/2014 12:05:47.
dietary fibre (DF). Their consumption has been associated with reductions in blood lipid
levels, triglycerides and low-density lipoprotein cholesterol, reduction in risk of
cardiovascular disease, increases in satiety, and decreases in blood glucose levels1, 2, 3. The
latter response is believed to be linked to the ability of DF to increase viscosity in the
upper gastrointestinal tract 4, 5. However, our earlier work demonstrated that following
simulated human digestion, the conditions and secretions present in the stomach and small
intestine (dilutions, pH changes and release of hydrolytic enzymes) led to profound
reductions in the viscosity of fibre-fortified solutions. Fibres consisting of more linear
structures, such as xanthan gum and guar gum, are more able to resist reductions in
viscosity in comparison to those fibres containing greater branching points and thereby
occupying a smaller hydrodynamic volume, such as soy soluble polysaccharide (SSPS).
The objectives of this study are to build on our earlier findings and explore the effects that
fibre-enriched solutions have on glucose modulation. More specifically, to determine the
effects, if any, that digesta viscosity may have on both glucose mobility and starch
hydrolysis inside a 2-stage simulated human digestion model mimicking both the gastric
and small intestinal environments.
2 EXPERIMENTAL
An assay was designed to observe the effect that a 2-stage simulated digestion has on the
viscosity of DF, when dissolved in solutions containing starch and protein (Table 1). Fibres
of varying structures (linear vs. branched) were dissolved in water at concentrations that
allowed for similar viscosities in a shear rate range of 25 - 100 s-1, as these are the shear
rates reported throughout digestion (Figure 1).
View Online
Figure 1: Viscosity profiles of solutions of 2% guar gum, 4% xanthan gum, 25% soy
soluble polysaccharide and 6.5% flaxseed gum, with starch (4%) and protein (6%)
An in vitro digestion method, mimicking gastric and small intestinal digestion, was
employed 6, 7. Briefly, 7.0 mL of gastric fluid [0.2%NaCl (w/w) in 0.7% HCl (w/v),
containing 3.2mg mL-1 pepsin] was added to 15 g samples, containing DF, starch and
sodium caseinate. Four glass balls were used to induce churning and agitation. The mixture
was incubated inside a shaking water bath at 37oC, at 175 rpm, for 1 h. Following
simulated gastric digestion, 4.6 mL bile fluid, containing 8 mg mL-1 bile salts, 14 mL
intestinal fluid, containing 5 mg mL-1 pancreatin dissolved in 0.5M sodium phosphate
buffer, and 2.9 mL amyloglucosidase (110 U/mL), were added to each solution. The
second stage of digestion, mimicking the small intestine, proceeded for 3 h. Despite
similarities in the initial flow behaviour of all solutions, rheological testing showed that
XG was capable of retaining a greater degree of its initial viscosity in comparison to the
remaining fibres (Figure 2).
An experiment was designed to observe the effects that the 4 fibre solutions have on
modulating glucose mobility inside a dialysis system. In vitro digestion proceeded as
View Online
Figure 2: Viscosity profiles of solutions of 2% guar gum, 4% xanthan gum, 25% soy
soluble polysaccharide and 6.5% flaxseed gum, with starch (4%) and protein (6%),
following 5 h in vitro digestion
described in Section 2.1, with the only modification being the transfer of digesta from the
flasks to dialysis tubes (MWCO 1000 Da) following the gastric step. Simulated small
intestinal digestion proceeded inside the dialysis tubes, which were tied off on both ends
and placed inside a solution containing 450 mL sodium phosphate buffer. 100 L aliquots
12/04/2014 12:05:47.
were withdrawn from the dialysate at time 0 (immediately after adding amyloglucosidase),
10, 20, 30, 40, 50, 60, 120, and 180 mins. The concentration of glucose inside the dialysate
was measured using the Megazyme D-glucose GOPOD enzyme assay kit (Megazyme
International Ireland Ltd., Wicklow, Ireland), according to the manufacturers instructions.
Throughout simulated small intestinal digestion, all fibres were able to lower glucose
mobility in comparison to the control (with no added fibre) (Figure 3). However, an
inverse relationship was observed between viscosity and glucose mobility whereby XG,
which was the most viscous fibre throughout digestion, was able to lower glucose transfer
across the dialysis membrane more effectively than the remaining fibres (p<0.05).
To determine whether the attenuating effects of DF are related to glucose diffusion alone, a
4% XG solution was prepared using D-glucose as a substitute for starch, thereby
eliminating any possible impact that XG may have on enzyme activity and starch
hydrolysis. The results indicate that the concentration of glucose in the dialysate, although
lower in comparison to the control, is higher when glucose is included in the initial
formulation (Figure 4).
This indicates that in our in vitro model, diffusion alone may not be the sole factor affected
by the addition of DF. Therefore, their ability to modulate glucose levels may be linked to
other physiological responses, such as impeding enzyme activity.
View Online
Figure 3: Available glucose (g/0.1mL) in the dialysate during 180 min in vitro small
intestinal digestion of control (no fibre), guar gum-, xanthan gum-, soy soluble
polysaccharide-, and flaxseed gum-fortified solutions
12/04/2014 12:05:47.
To ascertain whether incorporating soluble dietary fibres into solutions is slowing glucose
diffusion or instead lowering starch hydrolysis in our in vitro model, we modified a
method to analyse glucose concentrations in the digesta. Solutions containing 4% XG, 2%
GG and 25% SSPS were digested using the previously described method. Subsequently,
3.0 mL aliquots were withdrawn directly from the digesta at time 0 (after addition of
amyloglucosidase), 30, 60, 120, 180, 240 and 300 mins, and placed inside centrifuge tubes.
The digesta were mixed with 30 mL absolute ethanol and centrifuged at 4500 rpm for 15
View Online
Figure 5: Available glucose (g/0.1mL) in the digesta of control (no fibre), xanthan gum-,
soy soluble polysaccharide-, and guar gum-fortified solutions, during 4 h simulated small
intestinal digestion.
mins. The concentration of glucose inside the supernatant was determined using the
GOPOD enzyme assay kit. All fibres were able to reduce the concentration of glucose in
the digesta when compared to the control, with the less viscous SSPS demonstrating a
weaker capacity by which it was able to modulate the in vitro glycaemic response (Figure
12/04/2014 12:05:47.
5).
The degree of starch hydrolysis of the 4 solutions was calculated using the following
equation8:
(1)
Upon completion of simulated small intestinal digestion, the amount of starch that was
hydrolysed was reduced through the addition of soluble fibre. An inverse relationship
exists between digesta viscosity and starch hydrolysis, whereby the most viscous XG was
able to attenuate the extent of starch hydrolysis by greater than 97%, in comparison to the
control, which was observed to have lost 40% of its dry weight of starch (Table 2). This
finding is consistent with previous reports, which demonstrated that after 5 h of in vitro
digestion 40-50% of starch is hydrolysed and that after 24 h, the majority of starch was
broken down8.
Table 2: Amount of starch being hydrolysed after 6 h in vitro digestion of control, soy
soluble polysaccharide and xanthan gum-fortified solutions
3 CONCLUSION
The results of the present study show that the addition of soluble DF leads to reductions in
glucose concentrations, which was attributed to both a reduction in starch hydrolysis and a
reduction in glucose diffusion across the dialysis membrane. Despite reductions in
viscosity following simulated gastric and small intestinal digestion, all soluble fibres
employed in this study were capable of lowering the amount of glucose present, in
comparison to the control without fibre. XG, being the most resistant to digestion and
retaining a greater degree of its initial viscosity, was able to reduce glucose concentrations
more effectively than guar gum, SSPS and flaxseed gum; this illustrates that an inverse
relationship exists between viscosity and the amount of glucose that is liberated. Moreover,
the ability of soluble fibres to modulate glucose levels, in vitro, may be linked to the
viscosity that is developed inside the lumen. Starch hydrolysis, in vivo, occurs during
various stages of digestion. It is initiated inside the mouth, and the bolus travels through
the esophagus to the stomach, where gastric emptying allows for the material to arrive in
the small intestine. Here, the majority of starch hydrolysis takes place. The two primary
events that occur, which allow for glucose to be taken up by the blood, are hydrolysis in
the lumen by digestive enzymes secreted from the pancreas and diffusion of resulting
glucose monomers across the brush border membrane. An increase in viscosity through the
addition of viscous soluble fibres to the diet may impede both of these events. In the
present study, although diffusion kinetics were affected by the addition of DF, starch
hydrolysis was reduced substantially, indicating a plausible effect on enzyme activity. A
thickening of the luminal contents could cause reduced mixing, whereby the -1,4 bonds
12/04/2014 12:05:47.
of the starch molecule are obstructed and the access of -amylase enzymes to those bonds
is suppressed. The present study raises several pertinent questions pertaining to the
mechanism by which soluble dietary fibres are able to provide glycaemic control, which
are being addressed presently, and furthers our understanding of how DF reduces glucose
levels.
References
1 INTRODUCTION
Block copolymers are being introduced in food research as promising stabilisers able to
control the digestibility of lipid-based systems such as oil-in-water emulsions.1, 2 The
ability of these surfactants to modulate the lipid hydrolysis would enable the release of
hydrophobic compounds, already dissolved in the oil phase, into specific locations of the
gastrointestinal tract.3 This requires an improved understanding of the mechanisms
underlying lipid digestion. This process takes place mainly in the duodenum (upper small
12/04/2014 12:05:48.
intestine) and involves the action of different agents competing for the oil-water interface
of emulsified lipids.4 Among them, bile salts secreted by the gallbladder play a crucial role
preparing this interface for the adsorption of the enzyme pancreatic lipase. The enzymatic
breakdown of lipids starts only after the complex lipase-colipase is attached to a bile salt-
covered oil-water interface. In this sense, the lipolysis per se is largely an interfacial
reaction that can be affected by the presence of emulsifiers onto the oil-water interface,
either more surface active than bile salts and/or resistant to displacement by bile salts.5
In this regard, triblock copolymers of the family of Pluronics have been shown to
fulfil both conditions.6, 7 Pluronic molecules comprise a central hydrophobic moiety of
polypropylene oxide (PPO) which adsorbs onto the oil-water interface, flanked by two
lateral hydrophilic chains of polyethylene oxide (PEO) which protrude in the aqueous
phase. This brush-like adsorbed conformation seems to protect the lipid interface from the
adsorption of duodenal components. However, lipolysis might be also affected by events
taking place in the bulk that may have an impact on the interfacial properties related to
access for digestion. For that reason, the aim of this work was to analyse the bulk
interactions between Pluronics (F127 or F68) and a bile salt (sodium taurodeoxycholate,
NaTDC) in Pluronic-stabilised emulsions. Differential scanning calorimetry allowed the
evaluation of these interactions in the aqueous phase. Then, their influence onto the oil-
water interface was studied by interfacial tension measurements, looking at possible
competitive adsorption. Finally, further details of the microstructure of emulsions under
the action of bile salt were visualised with scanning electron microscopy. Results are
discussed comparing the molecular properties of Pluronic F127 and F68. Namely, F127
has longer PPO and PEO chains than F68, in such a way that PPO/PEO ratio is also higher
(Figure 1).
These new findings can be exploited in tailoring both, novel food and
pharmacological matrices with improved functional properties, while increasing the scope
for identifying functionalities of other potential (bio)polymers.
View Online
Pluronic F127 (MW 12600 g/mol) and F68 (MW 8400 g/mol), as well as the bile salt
NaTDC (MW 521.7 g/mol) were used without further purification. Highly refined olive oil
was purified to eliminate surface active impurities.7 All chemicals were purchased from
Sigma-Aldrich.
Pluronics and NaTDC were dissolved in 1.13 mM phosphate buffer (pH 7). Pluronic-
stabilised emulsions were prepared by mixing 10 wt/wt % olive oil with 90 wt/wt %
Pluronic solution at 1 wt/wt % with a high speed ULTRA-TURRAX homogenizer.
12/04/2014 12:05:48.
For the experiments in the presence of bile salt, 300 l of NaTDC solution was added
to 5 mL of Pluronic sample (either aqueous solution or emulsion) at room temperature. The
final bile salt concentration ranged from 0 to 50 mM whereas Pluronic concentration was
fixed either at 1 wt/wt % for DSC experiments or at 0.5 wt/wt % for interfacial tension
measurements. For the experiments in the presence of NaCl, the phosphate buffer
contained 0.15 M NaCl.
Interfacial tension measurements were carried out in a pendant drop tensiometer fully
assembled and developed at the University of Granada. The whole set-up is computer
controlled through the software DINATEN which fits the experimental drop profiles to
the Young-Laplace equation of capillarity by using Axisymmetric Drop Shape Analysis
(ADSA). As outputs the drop volume V, the interfacial tension J, and the interfacial area A
are provided. The adsorption process was recorded at constant interfacial area. The drop
was immersed in a glass cuvette (Hellma), which contained the oil phase and was kept at
View Online
constant temperature. The interfacial tension of the clean interface was measured before
every experiment obtaining values of (29.5 0.5) mN/m at 20 C.
A 3D SEM (FEI Quanta 200, Hillsboro, Oregon, USA), featuring a cryotransfer unit
(PP2000T, Quorum Technologies Ltd, Hailsham, UK) and an micromanipulator
(Omniprobe Inc., Dallas, Texas, USA), was used to visualise the internal structure of the
emulsions. It was operated at 15 kV and 3 1010 A. The sample was mounted on the
sample holder and plunged into slushy nitrogen and transferred into the evacuated
cryotransfer unit antechamber held at 140C. The sample was fractured using a cold knife
to expose the internal structure. Then, the freshly exposed emulsions surface was sputter-
coated with a thin conductive layer of platinum to allow high resolution imaging in the
SEM. The samples were finally transferred under vacuum from the antechamber into the
SEM chamber, also held at 140C.
The interactions between bile salt and Pluronics are first evaluated in aqueous phase before
considering the scenario of the bulk in Pluronic-stabilised emulsions. Figure 2 shows the
12/04/2014 12:05:48.
Figure 2: DSC traces on heating and on cooling of 1 wt/wt % Pluronic solutions in the
absence and presence of NaTDC at different concentrations.
Figure 3 shows the enthalpy values of the transition peaks on heating as a function of
the added bile salt concentration. The enthalpy values on cooling in the presence of bile
salt were again equal to those on heating, confirming the interaction phenomenon between
Pluronic and bile salt. It can be seen clearly that the enthalpy of Pluronic micellization is
12/04/2014 12:05:48.
reduced to zero at lower concentration of bile salt for F68 (10 mM NaTDC) than for F127
(around 50 mM NaTDC). Hence, it seems that the more hydrophobic F127 is able to
interact with larger amount of bile salt molecules.
Since bile salts are negatively charged at pH 7, the effect of NaCl at a physiological
concentration (0.15 M) was also studied. Figure 3 shows a similar trend in the enthalpy
values in the presence of NaCl when increasing the bile salt concentration. However, the
profiles for both Pluronics are slightly shifted to higher values of the enthalpy within the
whole range of NaTDC concentration. Eventually, the bile salt concentration needed to
completely dissolve the micellization of Pluronic is slightly higher in the presence of NaCl.
Since the charge of NaTDC is screened by increasing the ionic strength, it is likely the
compact arrangement of bile salt molecules, with a decreased repulsion between them, on
the hydrophobic portions of Pluronic. Therefore, at a certain NaTDC concentration, it
seems possible that there are more molecules of bile salt forming mixed micelles.
The interfacial tension values of Pluronic/NaTDC mixtures at the olive oil-water interface
are displayed in Figure 4, as a function of the bile salt concentration after 1 h of adsorption.
Pluronic concentration was fixed at 0.5 wt/wt %. Results are discussed comparing with the
interfacial behaviour of the bile salt in the absence of Pluronic. At low NaTDC
concentrations, a strong deviation of the curves of mixed systems from that of pure bile salt
is observed. Conversely, the curves of mixed systems approach that of pure bile salt at
high NaTDC concentrations. This indicates that the adsorption process of mixed systems
evolves from a Pluronic-controlled adsorption to a bile salt-controlled adsorption as
NaTDC concentration increases. Nevertheless, at the highest NaTDC concentration, the
mixture of Pluronic (either F127 or F68) and bile salt still decreases the interfacial tension
12/04/2014 12:05:48.
to a larger extent than the bile salt alone, suggesting the coexistence of both species onto
the interface. Hence, the results reveal the ability of Pluronics to compete for the available
interfacial area, even when the bile salt governs the process of adsorption, because the bile
salt is in excess with respect to Pluronic.
Some differences concerning both types of Pluronic are also observed. The mixture of
F68 and bile salt shows a maximum in the interfacial tension within the low NaTDC
concentration regime. This maximum might be attributed to the less surface activity of the
complexes formed by Pluronic and bile salts, as detected by thermal analysis, and as it was
previously reported for mixtures of F127 and ionic surfactants.13 This feature may not be
reflected in the mixed F127/bile salt system due to the lower interfacial tension provided
by the polymer. It can be better observed at lower concentrations of Pluronic.13 In addition,
the mixture of F68 and bile salt shows a larger drop in the interfacial tension at high
NaTDC concentrations, which suggests greater susceptibility of F68 to the presence of bile
salt competing for the oil-water interface. This might be explained by the lower surface
activity of F68 due to its lower hydrophobicity.7
When considering later on the scenario of emulsions stabilised by Pluronic, it must be
taken into account the fact that the surface of oil droplets are already covered by Pluronic.
Therefore, it might be even more difficult for bile salt to access the oil-water interface.
A similar set of DSC experiments were carried out for emulsion samples stabilised by
either 1 wt/wt % F127 or F68 before and after addition of bile salt at a fixed physiological
concentration of 10 mM NaTDC. DSC traces on heating are reported in Figure 5 only for
the F127 system, since they are analogous for F68-stabilised emulsion. Once more, the
results are reversible on cooling. Pluronic-stabilised emulsions display an endothermic
peak located within the same range of temperature as the micellization process shown by 1
12/04/2014 12:05:48.
wt/wt % Pluronic solutions. This suggests that the thermal event comes from the non-
adsorbed Pluronic remaining in the continuous phase. Nevertheless, the heat flow is lower
than that in Figure 2. This is probably due to the lower concentration found in the bulk, as
a large fraction of Pluronic is adsorbed on the surface of oil droplets.12 In order to check
the origin of this endothermic peak, the emulsion sample was centrifuged and the subnatant
(aqueous phase) was collected and analysed. Indeed, the subnatant reproduces the same
endothermic peak as the original emulsion sample, corroborating that the non-adsorbed
Pluronic mainly contributes to the DSC peak observed in Figure 5.
Then, the addition of bile salt to either emulsion or subnatant, gives rise to the
disappearance of the micellization peak. Hence, interactions between non-adsorbed
Pluronic and bile salt still take place in the aqueous phase of emulsions, hindering the
formation of Pluronic micelles and probably preventing bile salt adsorption on the surface
of oil droplets, as suggested by interfacial tension measurements at the oil-water interface.
Figure 6 shows the microstructure of these emulsions stabilised by either F127 or F68
in the absence and presence of the bile salt (10 mM NaTDC). In the absence of NaTDC,
lipid droplets with a relatively smooth surface are observed, surrounded by a homogeneous
aqueous phase. Interestingly, cryo-SEM reveals the presence of lumps in the aqueous
phase upon addition of NaTDC. The image of the system stabilised by F127 shows the
surface of two lipid droplets, whereas the image corresponding to the sample stabilised by
F68 illustrates the interface between a lipid droplet and the aqueous phase. It can be seen
how this granular texture also affects the surface of lipid droplets that now appear much
larger than in the absence of bile salt. Although Pluronics interact and compete for the oil-
water interface with bile salt and resist displacement by the bile salt, 6, 7 it is likely that the
bile salt co-adsorbs, penetrating the interfacial layer of Pluronic, hence turning it more
susceptible to some destabilization phenomena such as coalescence.2
12/04/2014 12:05:48.
4 CONCLUSIONS
The combination of bulk and interfacial events provides a new perspective into the
interaction of bile salts with Pluronics within emulsions systems. Pluronic seems to prevent
bile salt from adsorbing onto the surface of oil droplets in emulsions due to a combination
of complexation and competitive adsorption. The more hydrophobic and larger Pluronic
(F127) better competes for the oil-water interface and interacts with greater amount of bile
salt molecules. These findings are relevant to the rational design in the control of lipid
digestion or having specific physiological response of functional foods. Findings from this
study underscore the importance of accurately combining colloidal and interfacial
knowledge in order to promote reliable advances.
Acknowledgements
Authors thank the financial support from the European Communitys Seventh Framework
Program (FP7-PEOPLE-2012-IEF) under Grant Agreement No. 326581.
References
1 INTRODUCTION
This study has investigated the preparation of emulsions stabilised by two different classes
of biopolymer; Sugar beet pectin (SBP), a plant derived protein-polysaccharide complex,
12/04/2014 12:05:50.
and hydroxypropyl methylcellulose (HPMC), a surface active polymer. Both systems are
capable of forming monodisperse, highly stable emulsions. The interfacial and
emulsification behaviour of polymeric stabilisers such as HPMC have been widely
studied.1,2,3,4 Nevertheless, to our knowledge, it has not been used as a model for in-vitro
digestion studies up to now. Protein-polysaccharide complexes have previously been
investigated as possible systems to potentially control the digestion of lipids.5,6 Sugar beet
pectins are naturally occurring protein-polysaccharide mixtures, however the digestion of
SBP-stabilised emulsions has yet to be investigated. Previous work has demonstrated that
protein-stabilised interfaces can be displaced by bile salts. The focus of this work has been
to understand how hydrocolloid emulsifiers interact with bile salts, as these interactions are
possibly more complex than those observed in protein-stabilised interfaces.
Surface potential measurements are indicative of surface composition in emulsions and can
be used to compare the resistance of these two biopolymers to the adsorption of bile salts.
A physiologically relevant mixture of two common bile salts (BS) sodium taurocholate
NaTC and sodium glycodeoxycholate NaGDC (52.7:47.3 molar ratio of NaTC:NaGDC),
12/04/2014 12:05:50.
were used at a range of concentrations. Both bile salts are negatively charged amphiphilic
surface active molecules. From each emulsion, samples were prepared to a final
composition of 0.1 % (w/v) hydrocolloid and 2% (%w/v) olive oil. After incubation at
37 C for 1h, samples were exposed to increasing concentrations of BS (0.1mM to 10mM).
After 30 min, samples were diluted to an appropriate concentration to carry out surface
potential measurements, with a final composition of 2mM Bis-Tris, 10mM NaCl, and
either in the presence or absence of each samples corresponding BS concentrations. The -
potential (ZP) of the emulsions was determined using a Malvern Instruments Nano DS.
Figure 2 shows the surface potential data obtained for emulsions stabilised by HPMC
or SBP. Surface potential measurements show that the SBP-stabilised emulsions (Figure
8 7
0.1%HPMC
7 a 6 b 0.25% SBP
0.25% HPMC
6 0.5% HPMC 0.5% SBP
5
0.75% HPMC 0.75% SBP
5
Volume (%)
Volume (%)
1 1
0 0
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000
Diameter (um) Diameter (um)
Figure 1: Droplet size distributions for olive oil emulsions stabilized by various concentrations of (a)
HPMC and (b) SBP.
View Online
-20 -10
ZP (mV)
ZP (mV)
-30 -20
-40 -30
-50 -40
-60 -50
0.01 0.1 1 10 0.01 0.1 1 10
[bile salts] mM [bile salts] mM
Figure 2: Surface potential data obtained for emulsions stabilised by (a) 1% SBP (dilutions in {
buffer; z buffer+BS) and (b) 1% HPMC (dilutions in buffer; buffer+BS), in the presence of bile
salts.
2a) have a negative surface potential (~-40mV). As the BS concentration increases, little
difference is seen in the ZP of the SBP emulsions, when analysed either in the presence
(closed symbol) or absence (open symbol) of BS in the dilution buffer. Figure 2b shows
that the non-ionic polymer HPMC has a neutral surface potential (~-1mV). The same
behaviour to that of SBP is observed in the absence of BS in the dilution buffer (open
symbol). However, when the dilution buffer contained BS (closed symbol) a pronounced
12/04/2014 12:05:50.
decrease in the surface potential was observed, suggesting the composition of the HPMC
stabilised interface has been altered. The difference between the results (+/- BS) suggests a
possible reversible interaction between biopolymer and bile salts. HPMC has been shown
to display complex adsorption behaviour which influenced competition with adsorbed
proteins.1
Previous measurements (not shown) have investigated the displacement of milk
proteins from the olive oil-water interface with bile salts. The measured surface charge of
the emulsion still diminishes in the presence of increasing bile salt concentration, despite
the dilution step for analysis having been carried out with buffer, suggesting that in this
case BS-adsorption is irreversible, as seen previously during the displacement of
E-lactoglobulin and Tween 20.8 However, the data observed for the interaction between the
hydrocolloid and bile salts, does not allow us to draw any conclusions as to how the SBP
and BS interact. The lack of change in measured surface potential does not necessarily
mean that the BS are not affecting the interface, only that other surface techniques are
necessary to determine the behaviour of the adsorbed layer.
To further resolve the behaviour of the mixed interfaces, the interfacial pressure and
interfacial dilation modulus of an olive oil drop stabilised with HPMC or SBP, were
investigated in the presence of BS. These parameters were measured using the pendant
drop technique with a FTA200 pulsating drop tensiometer (First Ten Angstroms,
Portsmouth, VA), as described previously.7 Briefly, an olive oil drop is formed at the tip of
a J-shaped needle, in the presence of HPMC or SBP. The interfacial pressure ( = 0 -
), where 0 is the interfacial tension of the clean surface (27.50.5 mN/m for the olive oil-
water interface at 37C) of hydrocolloid solutions at the oil-water interface were recorded
for 50 min. After which BS were added sequentially to give final concentrations of 0.1mM,
View Online
1mM and 10 mM BS, and the dilatational response recorded for 15 min following each
addition. Analysis of the recorded images is carried out by fitting the experimental drop
profile to the YoungLaplace capillarity equation yielding the drop volume, interfacial
tension , and surface area.
The surface pressure as a function of time for SBP and HPMC solutions following
addition of BS are presented in Figure 3. Figure 3a shows the results for 0.1% SBP, higher
SBP concentrations were too turbid to allow an accurate fit of drop profiles. An increase in
surface pressure is observed on adsorption of the pectin at the interface from 0 to 14
mN/m after approximately 1h, SBP typically takes several hours to reach a steady state, as
would be expected considering its high molecular weight. If the bulk solution is rinsed
with pure water, the interfacial tension becomes stable and no significant subsequent
change is observed, confirming that the adsorption of the SBP is irreversible.9
In contrast HPMC adsorbs much more quickly, even at 0.001 % (data not shown),
reaching a pseudo-equilibrium within 1 minute. The values suggest surface activities
similar to those observed previously both at the air-water 1 and oil-water interface,10 and
suggest a critical aggregation concentration at the interface of around 0.001%. A much
higher concentration of SBP is required to saturate the interface as previously shown when
the bulk concentration of SBP reaches 2%.11 The different behaviour observed between
HPMC and SBP at the olive oil interface is due to their different structure and adsorption
mechanism as previously observed at the MCT interface.2 The saturation adsorption values
for HPMC suggest a closely packed interfacial layer. The areas occupied per molecule are
unusually small as compared with the molecular dimensions of HPMC, indicating that only
a few segments of the polymer chain are adsorbed at the interface as is described for
12/04/2014 12:05:50.
polymers by the trainlooptail model.10,12 However, the presence of pectin chains in SBP
prevents the proteins forming a tightly packed layer at the interface.13
The dilatational modulus of SBP adsorbed films increased with adsorption to E40
mN/m in around 1 h (Figure 4a). This value is comparable to that previously reported at air
water interface13 and slightly higher than that found for -lactoglobulin at the olive oil
interface.7 These results suggest that SBP forms an elastic film at the olive oil/water
interface, which is considered to arise from adsorption and association of the protein
component at the interface. The dilatational modulus of HPMC films adsorbed at the olive
oil-water interface hardly increases (Figure 4b) and is significantly lower than that seen to
occur for SBP films (Figure 4a). The lower value of the dilatational modulus attained by
HPMC films suggests that this biopolymer adopts a very different conformation at the
Figure 3 Surface pressure data as a function of time for the adsorption of (a) 0.1% SBP and (b) 0.1%
HPMC at an olive oil-water interface followed by the addition of bile salts (0.1mM; 1mM and 10 mM)
View Online
interface compare to the SBP. Figures 3 and 4 show how both the surface pressure and
elastic modulus respectively of SBP and HPMC networks are disrupted by the addition of
bile salts. Figure 3 demonstrates that as the concentration of bile salts was increased, the
surface pressure increased, showing that the bile salts are adsorbing to the interface. At low
concentration (0.1mM), the bile salts appear to be able to disrupt the interfacial SBP
network, as seen from the initial decrease in the dilatational modulus (Figure 4a). At 1mM
bile salt, the dilatational modulus continues decreasing with little change in surface
pressure. On the other hand the addition of 0.1mM and 1mM bile salt to HPMC, results in
a relatively smaller increase of surface pressure compared to that observed for the SBP at a
comparable bile salt concentration. The dilatational modulus data for HPMC following
addition of 0.1mM and 1mM bile salt also remains relatively unchanged with only slight
increase in surface pressure observed (Figure 4b).
However, the elastic modulus of the HPMC network is eventually diminished with
increasing bile salts concentration. At 10mM bile salt, a sharp increase in surface pressure
was observed with the surface pressure reaching a comparable value to those observed for
SBP. Similarly the dilatational moduli decreased towards zero for both the HPMC and SBP
interfaces. These results suggest that the bile salts are interacting with both HPMC and
SBP networks at the olive oil interface, but the mechanism of interaction is different for
each polysaccharide. The behaviour observed in the SBP network is comparable to that
observed for a pure protein7 probably due to the presence of a protein moiety in the SBP
structure.11 A similar behaviour was also observed in the displacement of -lactoglobulin
from a tetradecane interface by Tween 208 which is active over a surface pressure range
comparable to that observed for bile salts. Nevertheless, previous studies have shown that
12/04/2014 12:05:50.
the interfacial SBP film is more resistant to displacement by Tween 20 than a pure protein
film, probably caused by the formation of linkages between pectin chains at the interface.13
2.4 Effect of SBP and Bile Salts on the Interactions between Olive Oil Droplets.
Force spectroscopy experiments were made using an Asylum MFP-3D AFM (Asylum
Research, Santa Barbara, USA). The preparation of olive-oil droplets and their attachment
to the cantilevers (Nanoprobe (NP) silicon nitride levers, k 0.070.08 N m-1) and surfaces
has been described elsewhere.8,9 Force-distance data was acquired, at a constant relative
View Online
There is a 19% reduction in the slope of the linear region for the force-distance data
obtained between bare olive-oil drops and the drops+SBP. In addition to this observation it
can also be seen that if we compare the relative distances over which the force in each
system begins to rise above the zero force baseline (as determined by the cantilever
deflection), it is greater for the drops+SBP than bare drops and consequently, the change in
force per unit separation is smaller for the drops+SBP, suggesting that the drops+SBP are
more deformable. The largest change in the measured slope occurs following the
interaction between oil-droplets where the drops+SBP have been exposed to bile salts. This
causes a 64% reduction in the slope of the linear region of the constant compliance region
compared to that of the bare drops, and a concomitant increase in the distance over which
12/04/2014 12:05:50.
the force is measured. Hence the interfacial structure of the olive oil droplets clearly
becomes more deformable in the presence of bile salts. The droplet deformability is
determined by the internal pressure which varies according to Laplaces Law (eq.1).
(1)
Since the radius (r) is kept constant in this experiment, the surface tension () governs the
deformability of the droplets. It is well known that adsorption of surfactants reduces the
interfacial tension and therefore should give rise to a concomitant decrease in the Laplace
pressure of the droplet, making it more deformable. Hence, the data indicate that
View Online
adsorption of bile salts to the SBP network at the olive oil interface is taking place,
corroborating the results obtained from the interfacial tension measurements.
An in vitro digestion model was used to simulate the conditions found in the GI tract and
to investigate the interaction of SBP and HPMC with enzymes, bile salts and lipids in the
duodenal environment. In vitro duodenal digestions of 1% HPMC or 1% SBP stabilized-
emulsions were carried out using a Titrando 836 titration device (Metrohm Instruments),
by in situ titration of the digestion mixture as the lipolysis progresses. The emulsions were
diluted in 2 mM bis-tris buffer pH 7, which contained 0.15 M NaCl and 0.01 M CaCl2 to
mimic the ionic conditions in the duodenum. Lipolysis was performed in the presence of
9.7mM bile salt mixture and 100g pancreatin added from a 1mg/ml stock of porcine
pancreas (Sigma Aldrich). The free fatty acids released during lipolysis were continuously
neutralised with 0.05 N NaOH to maintain the pH of 7.7. The volume of NaOH required to
maintain a constant pH was used to calculate the concentration of titratable free fatty acids
released during lipolysis, which is proportional to the total fatty acid released.
Figure 6 shows the titratable fatty acids liberated during the course of lipolysis for
1%HPMC and 1%SBP stabilized-emulsions. Both the HPMC and SBP stabilized
emulsions have a comparable surface area: substrate ratio (10847 cm2/ml and 11086
cm2/ml respectively), It can be observed that a similar amount of fatty acids were produced
during their digestion, with the HPMC networks appearing slightly more resistant to
lipolysis than SBP under the in vitro duodenal conditions investigated. Further experiments
12/04/2014 12:05:50.
are required to test the premise established within this research work and to further explore
the possibility of controlling emulsion lipid digestion through interfacial modifications
such as the use of surface active polysaccharides.
3 CONCLUSIONS
This study forms one of the steps in the understanding of the interfacial mechanisms that
control the breakdown of food emulsions during digestion. The interaction of two
structurally different biopolymers, SBP and HPMC, with bile salts at the olive oil-water
interface has been analysed by combined use of surface potential measurements,
Figure 6 Titratable free fatty acid released from HPMC and SBP stabilised olive oil emulsions
under in vitro duodenal conditions
View Online
Acknowledgements
The authors thank Mike Ridout for advice and experimental assistance. This work was
12/04/2014 12:05:50.
supported through the EU, 7th Framework Programme, Marie Curie IEF, Grant no. 299920.
References
1 J-C. Arboleya and P.J. Wilde, Food Hydrocolloids, 2005, 19, 485.
2 X. Li, S. Al-Assafa, Y. Fanga and G.O. Phillips, Carbohydr. Polym., 2013, 91, 573.
3 R.N. Ziga,, O. Skurtys, F. Osorio, J.M. Aguilera and F. Pedreschi, Carbohydr.
Polym., 2012, 90, 1147.
4 N.A. Camino, C. Carrera Sanchez, J.M. Rodrguez Patino and A.M.R. Pilosof, Food
Hydrocolloids, 2012, 27, 464.
5 O.K. Simo, Y. Mao, T. Tokle, E.A. Decker and D.J. McClements, Food Res. Int.,
2012, 48, 337.
6 U. Lesmes and D.J. McClements, Food Hydrocolloids, 2012, 26, 221.
7 J. Maldonado-Valderrama, N.C. Woodward, A.P. Gunning, M.J. Ridout, F.A.
Husband, A.R. Mackie, V.J. Morris and P.J. Wilde, Langmuir, 2008, 24, 6759.
8 N.C. Woodward, A.P. Gunning, J. Maldonado-Valderrama, P.J. Wilde and V.J.
Morris, Langmuir, 2010, 26, 12560.
9 A. Gromer, R. Penfold, A.P. Gunning, A.R. Kirby and V.J. Morris, Soft Matter, 2010,
6, 3957.
10 C. Wollenweber, A.V. Makievski, R. Miller and R. Daniels, Colloids Surf. A:
Physicochem. Eng. Aspects, 2000, 172, 91.
11 C.M. Siew and P.A. Williams, J. Agric. Food Chem, 2008, 56, 4164.
12 J.Hunter, Foundations of Colloid Science, Clarendon Press, Oxford, 1986, vol.I, p. 49.
13 A. Gromer, A.R. Kirby, A.P. Gunning and V.J. Morris, Langmuir, 2009, 25, 8012.
SYNERGISTIC ROLES OF ALGINATES AND -GLUCANS IN GASTRIC RAFT
FORMULATIONS
1 INTRODUCTION
Heartburn and gastro-oesophageal reflux disease (GORD, GERD USA) are prevalent
among millions of people worldwide and these conditions can affect significantly their
quality of life.1,2 Most individuals with heartburn seek over-the-counter (OTC) and some-
times prescription only medicine (POM) treatments which include antacid formulations,
antacid/alginates, H2-receptor antagonists and proton-pump inhibitors. The direct medical
cost is over $12.2 billion annually and the indirect cost exceeds $81.7 billion in USA
alone.3
There are many antacid products used for the treatment of GORD.4-11 The antacid
only products neutralise gastric acid and relieve the discomfort from acid reflux. The
12/04/2014 12:05:52.
other types include the capacity for gastric raft formation whilst neutralising gastric acid.
Gastric rafts show better GORD relief than antacids alone.9 When alginates are used in
GORD formulations, a buoyant gel or raft forms on contact with the gastric acid due to the
formation of carbon dioxide from carbonates and bicarbonates within the polysaccharide
formulations. Divalent cations, usually calcium ions (from the added salts) are used to
bridge the alginate carboxylic acid groups together (discussed below). Alginate based
heartburn formulations represent major consumer health products and include brands such
as Gaviscon, Rennie and Algicon.5,10,11
Raft formation incorporates the result of several coordinated interactions and chemical
reactions within the stomach. Most rafting formulations contain sodium alginate, calcium
carbonate, sodium bicarbonate, plus colours, flavours and sweeteners. The exact
formulations and types of alginate used are brand specific. Alginates form a gel rapidly
when in contact with both gastric acid (acid gel) and divalent cations such as calcium ions
(calcium gel). Calcium gels are more robust than acid gels although with time in the
stomach, the calcium ions are progressively displaced by protons. In parallel, carbon
dioxide released from the carbonates by reaction with gastric acid (HCl) forms bubbles
which are trapped in the gel. This causes gel expansion by the trapped gas and as a
consequence the gel floats on the surface of the acid. Gastric retention is enhanced by the
small volume, the floating formulation and the structure although, peristalsis linked to
gastric emptying, progressively removes the rafting structure from the stomach. For long
lasting relief it is obviously desirable to retain the raft in the stomach as long as possible.
Alginates are composed of (14)--D-mannuronic acid (M) and -L-guluronic acid
(G) sugar units in the form of homopolymeric (-MM- or -GG- blocks) and heteropolymeric
sequences (-MG- or -GM- blocks). The -(14) polymeric linkages of guluronic acid res-
idues cause steric hindrance with respect to the carboxyl groups, whilst M residues create a
more flexible conformation in solution.12 The molecular basis for the greater rigidity of G-
blocks (compared to M) is discussed in detail elsewhere.13,14 High ratios of GG facilitate a
View Online
2.1 Materials
Alginates (Manugel LBA, and Protanal LFR 5/60) were sourced from FMC Biopolymer
(Ayr, UK), while -glucans (dextrin from tapioca starch, Crystal Tex 626; native waxy
maize starch, CFH-368 and pre-gelatinised maize starch, B37) were all sourced from
National Starch (Manchester, UK). Calcium carbonate (471-34-1) was obtained from
Mineral Technology (Birmingham, UK) and sodium bicarbonate (144-55-8) from Brunner
View Online
Mond and Co Ltd (Winnington, UK). The -amylase (A-6211, 43 U.mg-1 solid, from A.
oryzae) was obtained from Sigma-Aldrich (Gillingham, UK).
2.2 Methods
2.2.1 Raft preparation. Raft preparation followed the general method of Hampson et
al,5 which represents British Pharmacopoeia Method BP 2010, V(III), 2348. Each standard
combination of raft forming reagents contained 534 mg of sodium bicarbonate (carbon di-
oxide source) and 320 mg of calcium carbonate (carbon dioxide and calcium source). The
reagents representing one dose (i) fixed as 1000 mg alginate to which up to 25% pre-
gelatinised starch (PGS) or dextrin were added (w/w) or (ii) where up to 25% (w/w) of the
alginate was replaced by the PGS or dextrin were dissolved/dispersed in 20 ml deionised
water. They were then transferred via syringes into 250 ml beakers containing 150 ml of
pre-warmed (37C) 0.1 M hydrochloric acid (HCl) and L-shaped test probes (2.2.3) held at
37C in a water bath. The rafts were left to develop for 30 min before they were tested
(2.2.3).
2.2.2 Raft staining with iodine. Rafts formed as above were transferred to the petri
dishes and sectioned with a knife. Iodine solution (I2/KI, 0.02 N) was used to stain the raft
sections before photographing. The PGS generated a blue colour with iodine while dextrin
generated a red/brown colour.
2.2.3 Raft strength. After 30 min of raft development (2.2.1), the beaker containing
12/04/2014 12:05:52.
the formed raft was placed on the stand of a TA-XT2 Texture Analyser (Stable Micro Sys-
tem, Surrey, UK) with the L-shape wire probe connected.5 The Texture Analyser arm
pulled the probe up and through the raft vertically at a rate of 5 mm.s-1 over a 10 cm dis-
tance. The force (measured in grams) was recorded 50 counts per second (cps). The maxi-
mum force was used as the indicator of raft strength.
2.2.4 Raft volume and weight. The volume and weight of rafts for each dose of the
formulations were determined by the method of Kapadia and Mane.23
2.2.5 Raft durability. The raft durability testing methodology was adapted from the
method of Hampton et al.6 Rafts were formed (2.2.1) in 500 ml screw-capped plastic bot-
tles, each with a filter plate containing twelve 15 mm diameter holes at the bottom of the
bottle. After 30 min raft development, the rafts were weighed and then returned to the bot-
tles. The sealed bottles were placed sideways in a shaking water bath at 37C with a shak-
ing rate of 95 rpm and amplitude of 28 mm. Only raft fragments larger than 15 mm were
left on the filter plates after the shaking stopped and were weighed with the plates at vari-
ous time intervals. The plates containing the larger raft fragments were placed back in the
bottles and the same shaking and weighing procedure was repeated up to 60 min or until no
raft material was left in the filter plates. The raft weight expressed against time (area under
the curve, AUC) was used as a measure of the raft durability of each formulation.
2.2.6 Surface cohesion force. For the measurement of surface cohesion, doses of
each raft format were dissolved/dispersed in 40 ml water with a magnetic stirrer for 60 min
at room temperature. The liquid was then transferred to a 13.6 cm diameter petri dish and
placed on the stand of a TA-XT2 texture analyser. An aluminium plate probe with a diame-
ter of 10 cm was connected to the TA-XT2 and was positioned on the surface of the liquid.
View Online
2.2.7 Pushing force. Doses of raft formats were dissolved / dispersed in 60 ml water
with magnetic stirring for 60 min. The liquids were then transferred to 75 ml beakers and
placed on the stand of TA-XT2 Texture Analyser. A steel ball probe with diameter of 12
mm attached to the Analyser was placed 5 mm above the liquid surface. The rate of probe
movement was 10 mm.s-1 with a travel distance of 20 mm, and the resistant force during
the procedure was recorded at a rate of 100 cps. The maximum force required during the
process was used as the indicator of the pushing force (g).
3. RESULTS
The two alginates (Protanal LFR 5/60 and Manugel LBA, FMC Biopolymer) used for this
study possessed very low viscosities (300-700 mPa.s for 10% solutions), high G:M ratios
12/04/2014 12:05:52.
(67:33 versus 63:27, based on the manufacturers data) and produced rafts rapidly. Protanal
LFR 5/60 produced significantly higher raft strengths than Manugel LBA (15.5 1.6 ver-
sus 12.8 1.2 g respectively, P 0.05), when the alginate dosage was set as 1000 mg, as
employed by others.5 In order to compare raft strengths of alginate or alginate--glucan
formulations, the mean raft strength of 1000 mg doses without -glucan was set as 100%
and the relative raft strengths (RRS, %) of formulations including -glucan were obtained
by comparing the strength data with that of alginate alone. The amounts of calcium car-
bonate and sodium bicarbonate per dose were standardised as any salt variation impacts on
raft functionality.5,6
The interaction of alginate with -glucan was first verified by iodine staining where
PGS stained blue in the rafts and dextrin showed a red/brown colour. Hence the PGS and
dextrin were distributed effectively and homogeneously throughout the rafts.
When -glucans were added to constant amounts (1000 mg) of Protanal LFR 5/60 then
tested for raft strength, up to 15% PGS addition caused an average increase of 11% while
up to 20% dextrin increased raft strength by an average of 13% (Figure 1a). For Manugel
LBA (1000 mg) formulations, an addition of up to 20% PGS caused a 15% increase in raft
strength on average with a 19% increase for the same amount of dextrin (Figure 1b). Any
further increase of PGS or dextrin did not show further increases in raft strength for either
alginate.
A reduction of alginate below 1000 mg per dose led (unsurprisingly) to lower raft
strengths (Figures 1c and 1d). However, when a portion of Protanal LFR 5/60 was replaced
with PGS or dextrin, up to a maximum of 15%, this produced raft strengths similar or even
higher than the alginate alone (Figure 1c). Up to a maximum of 25% replacement of Ma-
nugel LBA with either PGS or dextrin also gave raft strengths equivalent to or slightly
higher than that of the alginate alone (Figure 1d).
View Online
Figure 1: The relative raft strengths of alginate and/or -glucan formulations. (a), raft
strengths of Protanal (P) LFR 5/60 (1000 mg) doses with the addition of pre-gelatinised
starch (PGS) or dextrin (Dex); (b), raft strengths of Manugel (M) LBA (1000 mg) doses
with the addition of PGS or dextrin; (c), raft strengths for doses where Protanal LFR 5/60
was replaced partially by PGS or dextrin; (d), raft strengths for doses where Manugel LBA
was replaced partially by pre-gelatinised starch or dextrin; (e), the effect of native starch
(NS) addition and substitution on raft strengths of Manugel LBA doses; and (f), the effect
of -amylase pre-treatments (0, 10, 60, or 300 s) on the raft strength of Manugel LBA - -
glucan doses.
As a control, native waxy maize starch was evaluated for its effect on raft strength
(Figure 1e). The starch in this form did not strengthen the rafts.
Raft volumes, weights and durability were also evaluated for typical (1000 mg) algi-
nate doses (Table 1). Here, both alginates produced similar raft volumes although Protanal
LFR 5/60 produced significantly higher raft weight and durability than Manugel LBA (P
0.05). The addition of PGS or dextrin did not produce obvious increases in the raft vol-
umes or weights for either alginate. The addition of dextrin caused a significant increase in
raft durability (P 0.05) while PGS did not. The addition of -glucan to compensate for
the reduction of alginate in the formulations (i.e. 1000 mg total polysaccharides) did not
change the raft weights although a slight decrease in volumes were observed.
View Online
The surface cohesive force between a liquid and solid surface may be used to mimic
the spreadability of liquid formulations on the tongue and oesophagus. Table 2 shows that
both the PGS and dextrin increased the force of the dosage systems. The PGS also caused
an increase in pushing force, while dextrin did not (Table 2).
When formulations containing 15% PGS or dextrin were treated with -amylase for 60
second (Figure 1f), raft strengths were still close to that of alginate (1000 mg.U-1) alone.
However, prolonged treatment with -amylase removed the strengthening effect due to
hydrolysis.
4 DISCUSSION
Gastric rafts are very effective for the treatment of heartburn and GORD. They do have
limitations, however, due to the chemistry of the alginates and the overall physiology of
the digestive tract. It could, for example, be advantageous to have a longer therapeutic du-
ration although the process of gastric emptying together with calcium-proton exchange and
loss of capacity to generate carbon dioxide in the stomach (due to salt depletion) limits op-
portunities.
Although the -glucans alone do not gel in a similar manner to alginates in gastric raft-
ing formulations (and would not be appropriate to be used alone for this purpose), they do
provide valuable components to be used within alginate based rafting matrices. Alginate
and -glucan physico-chemical and biological properties create synergies. These synergies
provide many potential therapeutic advantages within rafting matrices, e.g. drug delivery.
It is well known24,25 that -glucans can form inclusion complexes with certain molecules
including analgesics.26 The incorporation of the -glucans into the rafts may modify fla-
vour profiles, mask taste, control colours, provide rheological advantages with respect to
sensory (texture) properties, control drug release etc.; supporting utilisation of the rafts in
View Online
many different ways for therapeutic applications within the stomach and beyond. Non
therapeutic advantages include the protection of kelp beds and the lower costs of the
starchy molecules.
Using alginates, starch rich neem leaf powder and calcium chloride for making parti-
cles for the purpose of pesticide delivery, 20 researchers have reported that a hydrogen bond
stabilised polysaccharide matrix containing alginate and starch residues may be formed.
This is stabilised further by calcium ions bridging the alginate carboxylic acid residues.
This molecular arrangement assumes that all of the -glucans are accessible to interact
with the alginate molecules which would only be the case if they were solubilised and fully
dispersed. In reality, there will probably be some homo-polymer (alginate or starch) do-
mains dispersed together with hetero-polymer alginate-starch regions.
Drug delivery applications of alginate-starch blends have been discussed elsewhere.27
The mechanical properties of these matrices and their drug delivery characteristics are
modified by the incorporation of starch in the matrices. Sodium starch glycolate has also
been utilised with alginate to make drug delivery systems too with different properties to
alginate alone.28 Alginate-starch film have also been discussed elsewhere.29,30 Mucoad-
hesive alginate-starch matrices have been discussed for the purpose of therapy31 although
starch is sometimes utilised simply as a filler.32 Alginate-starch matrices have also been
used to encapsulate bacterial cells and prolong their viability.33
Where authors describe alginate-starch matrices for technical / pharmaceutical applications
such as drug delivery, little attention is paid to the source of the starch and often the nature
of the starch extent of gelatinisation, composition etc. is ignored. Many technical aspects
of these systems are thus not explored or understood although critical for the utilisation of
12/04/2014 12:05:52.
the matrices for specific delivery applications. These features will undoubtedly impact on
functionality as will the presence and amount of any other components such as the chemis-
try of specific drugs within the matrices.
5 CONCLUSION
The addition of -glucans to alginate raft formulations produced a strengthening effect and
partial replacement of alginate with -glucans produced equivalent raft strengths to that of
alginate alone. Dextrin showed a significantly greater impact on raft durability than PGS.
The addition of -glucans in the gastric raft formulations have other potential therapeutic
advantages, such as a reduction of bulking fibre boluses, possible carrier / controlled
release system for drugs incorporated into the rafts, textural improvement, cost of product
and reduction of the need to utilise kelp beds for the production of alginates.
References
1. J. Dent, H. B. El-Serag, M-A Wallander and S. Johansson, Gut, 2005, 54, 710.
2. G.N. Tytgat, R.C. Heading, S. Mller-Lissner, M.A. Kamm, J. Schlmerich, A.
Berstad, M. Fried, S. Chaussade, D. Jewell and A. Briggs, Aliment. Pharmacol.
Ther. 2003, 8, 291.
3. D.A. Peura, R.R. Berardi, J. Gonzalez and L. Brunetti, Pharm Therap., 2011, 36,
434.
4. K.G. Mandel, B.P. Daggy, B.P. Brodie and H.I. Jacoby, Aliment. Pharmacol. Ther.,
2000, 14, 669.
5. F.C. Hampson, A. Farndale, V. Strugala, J. Sykes, I.G. Jolliffe and P.W. Dettmar,
Int. J. Pharm., 2005, 294, 137.
6. F.C. Hampson, I.G. Jolliffe, A. Bakhtyari, G. Taylor, J. Sykes, L.M. Johnstone and
12/04/2014 12:05:52.
1 INTRODUCTION
Dysphagia is the medical term used to describe swallowing problems resulting from a
disorder in the mechanics of swallowing which can lead to impairment in the safety,
12/04/2014 12:05:54.
efficiency or quality of eating and drinking. Dysphagia can be caused by many disorders
including neurological disorders, stroke, traumatic brain injury, Huntingdons disease,
multiple sclerosis, Parkinsons disease and cerebral palsy. Many of these conditions are
associated with the elderly and, with the continuing demographic shift, dysphagia is a
growing problem. The responsibility for the management of the condition in the elderly is
not only in the hands of health professionals but also with individuals and carers in the
home.
There seems to be little doubt among health professionals that thicker foods are easier and
safer to swallow than thinner foods however the problem remains of how do you define
thick. This problem is further complicated by the variety of agents used to thicken the
foods. Several studies have tried to implement simple consistency and/or viscosity
measurements to help in the preparation of thickened foods for patients.1 7 Table 1
summarises the various recommendations made and shows that there is considerable
variability in both subjective consistency descriptors and conditions for objective
measurements. The most widely used classification appears to be that proposed by the
National Dysphagia Diet project. This was based on a study in which 10 orange drinks,
prepared with viscosities, (measured at 11 s-1), ranging from 10 mPa.s to 3000 mPa.s by
addition of modified starch, were presented to a number of Speech Language Pathologist
(SLPs) and Registered Dieticians (RDs). They were asked to write down what they
considered to be the best label for the liquid. They could stir the drinks but could not place
the samples in their mouths. The study concluded that there was very poor agreement in
the labels given. For example one sample was given a total of 51 different labels.7 The
most common labels from this study were adopted by the National Dysphagia Diet and
accepted by manufacturers of commercial thickeners. Corresponding viscosity ranges,
measured at a shear rate of 50s-1, were also given to each category (Table 2).
View Online
It is well known that the rheological properties of thickening agents vary considerably8.
Starch has been widely used in instant thickening products for dysphagia but more recently
guar gum and, in particular, xanthan gum is becoming increasingly popular. A shear rate
of 50s-1 is often chosen as being representative of the shear rate in the mouth.9 This is,
however an over simplification as it has been shown subsequently that the effective shear
rate at which samples of similar perceived viscosity, had the same measured viscosity,
decreased as viscosity increased.10 Further more the perceived thickness of xanthan gum
solutions and food products with weak gel properties were underestimated.11 In this case it
was concluded that dynamic viscosity measurements provided a better objective
measurement of fluid texture than conventional rotational measurements.
These studies were concerned with perceived thickness in the mouth and the correlations
found are probably most relevant to the oral preparatory stage of swallowing during which
the bolus is formed and assessed in terms of its suitability for swallowing. The rheological
properties of the bolus significantly influence the swallowing process and the use of
hydrocolloids to control the rheology can greatly help in the management of the condition.
Subjective Objective
Instrument Reference
consistency descriptors consistency measurement
Nectar 14 1cm per 30s
Bostwick
Honey 8 1cm per 30s 1
consistometer
Pudding 4 1cm per 30s
12/04/2014 12:05:54.
1 9 mPa.s
10 99 mPa.s Viscosity at
None 5
100 9999 mPa.s 50 s-1
> 10000 mPa.s
> 4 cm
3.0 3.9 cm Line spread
None 5
1.1 2.9 cm test
< 1 cm
Syrup
6
Yoghurt
Thin 1 10 mPa.s
Nectar-like 70 125 mPa.s Viscosity at
7
Honey-like 450 550 mPa.s 11 s-1
Spoon-thick > 2750 mPa.s
Factors such as size, shape, volume, consistency, pH and temperature are important in a
persons evaluation of the readiness of the bolus for swallowing12. Once swallowed, other
factors may be important and it is likely that the different stages of swallowing will require
different rheological measurements to obtain relevant correlations. Exactly what these are
remains to be established.
Recently in Japan proposals for classifying thickness of products for dysphagia, known as
Toromi, have been made based on viscosity at 50s-1 and the Line Spread Test (Table 3)12.
The LST test consists of placing the Toromi solution (around 20mL) in a metallic cylinder
of 3 cm diameter and 2.8 cm height which is placed on a graduated sheet (Picture 1). The
cylinder is then removed and the diameter (or the radius) of the spread is measured after a
set time.
The LST has a number of practical advantages over the viscosity since it is very simple to
perform, does not require viscometers or rheometers and can be made with a minimum of
training.
In this study five commercial products for dysphagia containing different hydrocolloids
have been tested according to the two tests currently proposed for use in Japan. Viscosity
was measured at 50s-1 as a function of concentration and compared to the LST results
measured over the same concentration range. The results have been compared and
discussed in terms of the rheological properties of the different thickeners used in the
products.
12/04/2014 12:05:54.
The commercial products used are listed in Table 4. Samples were prepared at
concentrations between 0.5% and 3%. To ensure complete hydration of the product and
eliminate any effects due to differences in dispersibility and or speed of hydration the
samples have been prepared according to the following procedure:
a. Weigh the corresponding amount of deionised water into a 250ml low shape beaker.
b. Position a 55 mm diameter propeller 1mm from the bottom, fix the stirring speed at
500rpm.
c. Pour the corresponding amount of powder instantly in 1 sec without stirring.
d. Immediately start stirring at 500rpm for 30 min.
e. Cover with parafilm and stir overnight with a magnetic stirrer.
f. Next day, make the LST test (average of 3 determinations) as well as the rheological
measurements.
Viscosity measurements were made at 20C at a shear rate of 50s-1 using a controlled stress
rheometer (Anton Paar MCR501) fitted with a 50 mm diameter cone with a gap of 209 m
and an angle of 2.005. The viscoelastic properties were measured at 20C with the same
rheometer. All measurements were made within their respective linear viscoelastic
domains and samples were fully hydrated before measurements were made.
The LST is made by pouring theToromi solution (around 20mL) into a metallic cylinder of
3 cm diameter and 2.8 cm height, placed on a graduated sheet (Picture 1). The cylinder is
12/04/2014 12:05:54.
then removed and the diameter of the spread is measured after 30s. The result is based on
an average of three measurements.
The relationship between thickener concentration and viscosity at 50s-1 is shown in Figure
1. It can be seen that the concentration dependence of viscosity is similar for all the
xanthan types while the starch type (Trommelin Gra) and the starch & guar type (Oillio)
show a different dependence. This is linked to the differences in the ability of each
hydrocolloid to thicken the water and their solution rheology. For example, xanthan and
guar gum are able to create viscosity at much lower concentrations than starch whereas
starch becomes much more effective at high concentrations. The concentration ranges
proposed in Table 3 fit well for the xanthan based products but not for the other thickener
types. For example the starch based product (Tromelin Gra) would require between 2
2.5% to meet the Mildly thick classification and dosing the Oillio product at 2 3%
would create a viscosity above the range for Extremely thick.
Name Type
Tromelin Gra Dextrin, Starch
Softia Dextrin, Xanthan
Oillio Dextrin, Starch, Guar
Quickly Dextrin, Xanthan, Calcium lactat, Na-Citrate
GRINDSTED
Dextrin, Xanthan
Xanthan NTJ-1
View Online
Figure 2 shows the LST as a function of the concentration. This test shows less difference
between the thickener types than the viscosity at 50s-1 with only the starch type Tromelin
Gra, being on a different curve. It shows that the starch based product spreads further than
the other products at the same concentration.
Figure 2: Concentration dependence of the LST (diameter) for 5 different Toromi types at
20C measured after full hydration in deionised water.
View Online
Comparing this data with Figure 1 draws one to the obvious conclusion that the lower the
viscosity at 50s-1 the greater the product will spread in the LST. However a plot of
viscosity at 50s-1 as a function of the LST diameter (Figure 3) does not provide a universal
correlation. In fact the correlation splits into two groups, one for the xanthan products and
one for the starch and starch/guar products. The LST value tends to be lower for the
xanthan based products compared with the starch and guar based products at equivalent
viscosity at 50 s-1. In other words, for the same viscosity, the starch type spreads more than
the xanthan type. This suggests that different rheological phenomena are involved in the
two tests and that the universal classification and concentration recommendations shown in
Table 3 may not be relevant for all the thickener types. A classification per thickener type
might be more appropriate.
Solutions were prepared at the recommended viscosities in Table 3 and the LST value was
measured. Indeed, when the actual LST values are measured at the viscosity at 50s-1 from
the classification in Table 3, it can be seen that only the pure xanthan types match the
recommendations for both viscosity and LST (Table 5). The deviation of the Quikly
sample which is also pure xanthan may be due to the presence of additional salts which are
known to impact the rheological properties of xanthan gum.
12/04/2014 12:05:54.
Figure 3: Viscosity at 50s-1 as a function of the LST value (biggest diameter of the spread)
measured after full hydration in deionised water.
View Online
Table 5: Classification according to Table 3 (LST values refer here to the radius of the
spread in mm).
To try to understand the rheological phenomena involved in the LST the viscoelastic
properties of each solution at the six concentrations were determined.
The complex viscosity, the complex modulus and Tan Delta were measured as a function
of the frequency. The figure below (Figure 4) shows the results for a xanthan type
thickener (Softia) and for a starch & guar type (Oillio) at the same concentration (3%).
The complex viscosity * can be seen as the viscoelastic flow resistance of the material.
12/04/2014 12:05:54.
The Newtonian plateau at the low frequencies for Oillio indicates the behavior of a
viscoelastic liquid at rest while * is rising towards an infinitely high value for Softia with
a constant slope approaching -1 attesting of a gel character as expected with xanthan.
This is confirmed with Tan Delta which is always below 1 for Softia (G>G) which is
typical of a viscoelastic gel while it is above 1 at the low frequencies for Oillio which
highlights a viscoelastic liquid (G>G). The complex modulus G* shows greater
frequency dependence for Oillio compared with Softia, confirming the viscoelastic liquid
property.
Figures 5 and 6 plot the complex modulus G* and the complex viscosity * as a function
of the LST results at high and low frequency respectively.
Figure 4: Complex viscosity (Pa.s), (), Complex modulus (Pa), () and Tan Delta (S)
as function of the frequency for: a) Xanthan based thickener (Softia) and b) Starch &
guar based thickener (Oillio) at the same concentration (3%).
View Online
Figure 5: Complex modulus and complex viscosity at high frequency (15.8 rad.s-1) as a
function of the LST diameter.
12/04/2014 12:05:54.
Figure 6: Complex modulus and complex viscosity at low frequency (0.125 rad.s-1) as a
function of the LST diameter (mm).
Both with the complex viscosity * and with the complex modulus G*, it seems that at low
frequency the two groups (xanthan versus starch type) are less obvious than at high
frequency especially at the low LST values (high viscosity). This indicates that the extent
of spread in the LST is not only governed by rotational viscosity but also by the dynamic
viscoelastic properties of the fluids particularly at low frequency. As discussed by
Morris11 viscosity at 50s-1 may underestimate the mouthfeel of products with weak gel
properties such as xanthan. This can explain why the LST values are lower for xanthan
type products at equivalent viscosity compared with starch (Figure 3).
These results show that the difference in viscoelastic properties between the thickeners is
an extra parameter to take into account when classifying the thickener type in a standard
classification. It would be judicious, when comparing the thickeners, to be under such
conditions so that the difference in viscoelastic properties is taken into account.
View Online
During the LST test the sample being only submitted to gravity, it is wise to say that only
the low frequencies are relevant to be considered here. And looking at the graphs above
and in particular at the Figure 6, it is clear that the fact that the relationship between G*
and LST at low frequency does not distinguish the 2 groups (i.e. in this area the
viscoelastic properties match) makes the LST be more adapted than the viscosity at 50s-1 to
12/04/2014 12:05:54.
classify the type of thickeners and it should be preferred towards the viscosity at 50s -1 to
make a universal classification.
If the viscosity at 50s-1 function of the LST (Figure 3) gives two clear groups, the relation
between the low shear viscosity (at 0.1s-1) and the LST (Figure 7) on the contrary gives a
unique correlation that confirms the shear rate of 50s-1 may not be the best to differentiate
the thickeners type .
4 CONCLUSIONS
The introduction in Japan of a new test, the Line Spread Test (LST) to quickly evaluate the
consistency of food thickeners has challenged the existing classifications and in particular
the use of viscosity at 50 s-1 to provide classification of the fluid consistency.
This work shows clearly that it is difficult to compare thickener systems with different tests
due to the rheological phenomena governing the test and the different thickening behaviour
of the hydrocolloids used in products for dysphagia. It is also difficult to make universal
dosage recommendations for different thickener types even if only one test is used due to
the differences in the concentration response between thickener types. Toromi
manufacturers would need to establish the corresponding values and concentrations for
each test specifically for their product.
The LST appears more adapted than the viscosity at 50s-1 to be able to provide a universal
classification of the fluid consistency since it allows a measurement where the viscoelastic
properties are taken into account for the different thickener types.
Even if the LST can help build a universal classification, the question as to which
rheological profile is the most appropriate for the management of dysphagia (i.e. xanthan
or starch based products) is one that remains to be answered.
View Online
References
England (2007).
9 F.W. Wood. SCI Monograph No. 27: Rheology and texture of foodstuffs (1968),
40.
10 F. Shama and P. Sherman. Identification of stimuli controlling the sensory
evaluation of viscosity. II. Oral methods. J Text. Stud., (1973) 4, 111.
11 E.R. Morris. Rheology of hydrocolloids. In, Gums and Stabilisers for the Food
Industry 2. (Phillips G.O., Wedlock D.J. and Williams P.A., Eds) (1984), 57.
12 Y. Yamagata, A. Izumi, F. Egashira, K-I Miyamota and J. Kayashita.
Determination of a suitable shear rate for thickened liquids easy for the elderly to
swallow. (2012) Food Science and Technology Research 18, 363
INVESTIGATION OF PHYSICO-CHEMICAL PROPERTIES OF GELATIN
MATRICES IN CORRELATION WITH DISSOLUTION STUDIES
1 INTRODUCTION
Gelatin is a versatile biopolymer obtained from a partial hydrolysis of collagen. The parent
collagen molecules consist of three single stranded alpha chains organized in a triple
helical structure. The triple helix structure is favored due to a unique amino acid
composition of the alpha chains, with a general amino acid sequence of: Glycine-X-Y,
where X and Y are usually proline (Pro) and hydroxyproline (Hyp), respectively. The
collagen molecules are organized into fibrils, which again are assembled into fibers. In
12/04/2014 12:05:57.
order to produce gelatin, the inter- and intra-molecular bonds within and between the
collagen molecules need to be broken. This is achieved by either acid or alkali
pretreatment, which give two types of end products: type A and B gelatin, respectively.
During the pretreatment with alkali, the amino acids glutamine and asparagine are
transformed into glutamic and aspartic acid. This change in composition gives the type B
gelatin a lower isoelectric point (4.9 5.2) compared to the type A gelatin (8 9). After
the pretreatment of the raw material, the gelatin is extracted by the use of hot water or
diluted acid. The dissolved fraction after this step is referred to as gelatin.1, 2
Gelatin is commonly used in formulation of pharmaceuticals and nutraceuticals, by
encapsulating the active ingredient into gelatin-based soft or hard gel capsules. In addition,
a new type of gelatin based formulation has been proposed in the form of an emulsified
gel.3 For all these applications the dissolution profile of the matrix in the gastrointestinal
tract (GIT) is of crucial importance. The dissolution of a gelatin matrix in the GIT may be
seen as a two-step procedure: gel-sol transition of the gelatin network, followed by
diffusion of the peptide strands from the concentrated matrix to the bulk liquid. Thus, it is
clear that the melting temperature may be of crucial importance for the investigation of
gelatin gel dissolution. The melting temperature is known for gelatin to be highly
dependent on the fraction of Pro and Hyp on the individual chains. Gelatin is mainly
produced from either bovine or porcine sources, in which the amount of Pro and Hyp is
almost constant. Besides the fraction of imino acids, the molecular weight distribution and
concentration of the gelatin, thermal history, pH and additives are all known to influence
the melting temperature of the gel.4 During hydrolysis and extraction of the collagen, the
secondary structure is destroyed as well as parts of the primary structure. Thus, depending
on the raw material and extraction conditions, gelatins with wide variations in molecular
size distribution, thereby also varying rheological properties/melting temperature, may be
obtained.
View Online
Dissolution studies are generally time consuming and it would hence be advantageous
to be able to predict the dissolution profile without performing in vitro measurements. The
scope of this work was to evaluate if any of the physico-chemical properties of the gelatin
matrices can be used to predict the dissolution profiles of the gelatin gels.
2.1 Materials
Type A and B gelatins with reported Bloom values of 160 g, 200 g and 260 g were kindly
supplied by Gelita (Mannheim, Germany) and used without further processing.
Information about the different gelatins is listed in table 1. All experiments were performed
using deionised water (MQ-water). HCl (37 wt.% HCl) was supplied by Merck
(Darmstadt, Germany). Acetaminophen was purchased from Sigma Aldrich (St. Louis.,
Missouri, USA).
sample was further mixed with an eluting buffer (0.1 M Na2SO4, 0.02 M Na2EDTA, 0.5 M
Trizma base, pH-adjusted to pH 9) to a ratio of 1:1. The mixture was filtered with a syringe
filter (1 m, glass fiber membrane, Pall, New York, USA) and stored at 4 C prior to
analysis. The size exclusion chromatography was performed at 40 C on two packed
columns (TSKgel G-6000 + 5000PWXL, Tosoh Bioscience, Tokyo, Japan) connected to a
Dawn DSP multi-angle laser light scattering photometer (Wyatt technology Corp, Santa
Barbara, CA, USA) followed by a Optilab DSP differential refractometer, P-10 cell (Wyatt
Technology Corp.). The data was analysed using the respective software (Astra, v6.1.1.17).
The weight average molecular weight (Mw, equation 1) and number average molecular
weight (Mn, equation 2) were determined from the molecular weight distributions of the
gelatins. In the equations, Ci represents the concentration (g/L), ni is the number of
molecules (mol) and Mw is the molecular weight (g/mol) of the individual molecules. The
polydispersity index (PI) is given as the ratio between Mw and Mn.
Table 1: Information about the different gelatins used in this study, with reported Bloom
strength given by Gelita. The abbreviation listed is used throughout this work.
n M i
2
i C M i i
Mw = i=1
n
i=1
n
(1)
n M
i=1
i i C
i=1
i
n n
n M i i C i
Mn = i=1n i=1
(2)
n
Ci
ni
i=1
i=1 M i
Acetaminophen was used as a marker to determine the dissolution profiles of gelatin gels
in vitro. It was a priori assumed that the release of acetaminophen from the gelatin gels
corresponded to the dissolution of the gel matrix. Gelatin gels were prepared by dissolving
gelatin (25 wt.%) into MQ-water at 55 C. Acetaminophen (0.906 wt.%) was dissolved
into the gelatin solution and the sample was degassed in a vacuum chamber (Nalge
company, Rochester, New York, USA) connected to a vacuum pump (Diaphragm Vacuum
Pump, Wertheim, Germany). The gelatin solution was gelled into small petri dishes (3.5
cm in diameter, 5.00 gram gelatin solution/dish) and two layers of Parafilm were used for
proper sealing. The gelatin gels were stored for 24 hours at 20 C and 60 %RH prior to the
12/04/2014 12:05:57.
dissolution studies.
evaporation, the applied sample was covered with silicone oil (10cS fluid, Dow Corning,
UK) prior to measurement. The viscoelastic properties of the sample were obtained by
using a temperature gradient (2 C /min), with a start and end temperature at 60 C and a
holding time of 15 minutes at 20 C. The gelling (Tg) and melting temperature (Tm) of the
gelatin gels were estimated as the temperature at which the phase angle corresponded to
45 in the cooling and heating process, respectively. G (Pa) was determined as the last
measuring point after curing at 20 C.
The molecular weight distributions (figure 1) for the different gelatins were determined by
SEC-MALLS measurements. From these measurements, the weight average molecular
weight (Mw), number average molecular weight (Mn) and polydispersity index (PI) were
determined (table 2).
As seen in figure 1, the different gelatins have varying amounts of high molecular
(HM>50 kDa) and low molecular (LM<50 kDa) peptide chains. This is highlighted by
investigating Mn and Mw seen in table 2. A higher Mn and Mw are measured with
increasing Bloom for both gelatin types due to a higher fraction of HM-chains. The
molecular weight distribution of the 160, 200 and 260 bloom, type A and B gelatin clearly
shows a shift towards higher molecular weights for the alkaline treated samples. It has
12/04/2014 12:05:57.
been proposed that acid pre-treatment causes hydrolysis mainly in the polar regions of the
peptide chain, while alkaline hydrolysis is less specific.6 A more extensive hydrolysis of
imino-rich zones (apolar) may hence occur, which gives a gelatin with a reduced gelling
potential compared to an acid treated sample. This may explain why a higher average
molecular weight of the alkaline treated gelatin is required in order to obtain a similar
bloom value.
1.8E-05
Differential weight fraction (mol/g)
1.6E-05
1.4E-05
1.2E-05 Type A - 260 bloom
Type A - 200 bloom
1.0E-05
Type A - 160 bloom
8.0E-06
Type B - 260 bloom
6.0E-06
Type B - 200 bloom
4.0E-06 Type B - 160 bloom
2.0E-06
0.0E+00
10 100 1000
kDa
Table 2 Determined average molecular weight (Mn and Mw) and polydispersity index (PI)
for the different gelatin samples. Mn ,Mw and PI being the average value of two replicates.
Gelatin Mn Mw PI
Type A - 260 Bloom 88.7 191.3 2.16
Type A - 200 Bloom 59.4 148.2 2.50
Type A - 160 Bloom 36.1 90.9 2.60
Type B - 260 Bloom 118.8 217.2 1.86
Type B - 200 Bloom 73.0 174.6 2.31
Type B - 160 Bloom 58.1 161.1 2.77
Large molecular chains (>1000 kDa) were observed for all the different gelatins
except for the type A, 160 Bloom gelatin. This gives the weight average a large value due
to these chains being emphasized in the calculation. The presence of such chains will
further influence the calculated PI, as seen in table 2, resulting in a high value. The present
result is in accordance with earlier studies, which have reported the presence of high
molecular weight fraction (>1000 kDa) for gelatin samples.2, 7
In order to investigate the rheological properties of the gelatin gels (25 wt.%), small strain
oscillatory measurements were performed. In figure 2, G (after 15 minutes of curing at 20
C) is shown. The melting and gelling temperatures are further given in figure 3.
12/04/2014 12:05:57.
25000
20000
15000
G'(Pa)
10000
5000
0
160A 160B 200A 200B 260A 260B
Figure 2 Measured G (Pa) after 15 minutes of curing at 20 C for the type A and B gelatin
gels (25 wt.% gelatin) with varying Bloom (160, 200, 260 g) determined by small strain
oscillatory measurements during a cooling-heating process from 60 C 20 C (15 of
minutes of curing) 60 C with a temperature gradient of 2 C/min. Three replicates were
performed on each gel.
View Online
40
38
36
34
Tm/Tg ( C)
32 Tg
30 Tm
28
26
24
160A 160B 200A 200B 260A 260B
Figure 3 Measured melting (Tm) and gelling temperature (Tg) for the type A and B gelatin
gels (25 wt.% gelatin) with varying Bloom (160, 200, 260 g) determined by small strain
oscillatory measurements during a cooling-heating process from 60 C 20 C (15 of
minutes of curing) 60 C with a temperature gradient of 2 C/min. Three replicates were
performed on each gel.
From figure 2, a clear correlation between the storage modulus and reported Bloom
values is observed. This was expected, since both terms represent elastic response. A
12/04/2014 12:05:57.
similar correlation is not observed between Bloom and melting/gelling temperature for the
different gelatins. An increase in Tm and Tg was however measured for the 260 Bloom
gelatins, with especially the type A gelatin being significantly higher compared to the other
gelatin gels.
Earlier, a correlation between the weight average molecular weight and gel firmness
for acid pretreated fish gelatins has been suggested.8 This is investigated for the gelatins
used in this study by presenting the weight (figure 4.A) and number average molecular
weight (figure 4.B) as a function of the Bloom values for the different gelatins. As seen in
figure 4, a correlation between Bloom and the weight/number average is observed for both
the type A and B gelatins used in this study. However, the trend is more noticeable for the
300 300
250 250
Bloom value (g)
Bloom value (g)
200 200
150 150
100 100
0 100 200 300 0 50 100 150
Mw (kDa) Mn (kDa)
4.A Type A Type B 4.B Type A Type B
Figure 4: The Bloom values (reported by Gelita) plotted against weight (figure 4.A) and
number average molecular weight (figure 4.B) for the type A and B gelatins.
View Online
In figure 5 the measured dissolution profiles of the different gelatin gels (25 wt.% gelatin)
is shown.
As seen in figure 5, a delayed dissolution is observed for the 260 Bloom gelatin gels,
especially for the type A gelatin, with only 50% of the gelatin matrix dissolved after 14
minutes. These results suggest that the correlation between the dissolution and the gel
firmness is not clear cut, since such a large difference in the dissolution profile is observed
for the 260 Bloom gelatin gels. However, there seems to be a correlation between the
12/04/2014 12:05:57.
amount of dissolved matrix and the measured gelling/melting temperature. The 260 Bloom
gelatin gels exhibited the slowest dissolution rate and also had the highest gelling and
melting temperatures, with the type A gelatin having both the slowest rate of dissolution
120%
Dissolved gelatin matrix
100%
80%
60%
40%
20%
Time (min)
0%
0 160A 10 160B 200A
20 200B 30 260A 260B
40 50
Figure 5: Dissolution profiles of type A and B gelatin gels (25 wt.% gelatin) with varying
Bloom (160, 200 and 260 g). The dissolution profiles were recorded in vitro using a
standard pharmaceutical dissolution unit and procedure (0.1 M HCl, 37 C, 75 RPM).
Five replicates were performed on each gel, with each replicate being the average value of
three individual measurements.
View Online
and highest Tg and Tm. The 260A and B gelatin gels were also the only ones with a
melting temperature higher than 37 C with the applied temperature gradient, hence
exhibiting solid-like behavior above the temperature of the artificial gastric fluid. Thus, it
would be expected that these gels would maintain as a solid for a longer time period after
exposure to the dissolution fluid and thereby showing a slower dissolution rate. The lower
Bloom gelatin gels dissolved quickly, at roughly equal rates. This may suggest that the
dissolution rate of the matrix is independent of the physical properties of the gel below a
certain melting temperature.
4 CONCLUSION
The present results suggest a correlation between the gelling and melting temperatures of
the gelatin gels and their dissolution profiles and that the information given by the Bloom-
value/storage modulus of the gels is not sufficient. Small strain oscillatory measurements
may hence be a useful tool in order to easier predict the dissolution profile of gelatin gels.
Information regarding the dynamic viscosity at 37 C may provide additional information
suitable for predicting the dissolution rate of the gels. In addition, studies regarding the link
between physical properties and dissolution profiles may potentially be used to prepare
tailor-made gelatin gels with predefined dissolution rates.
Acknowledgements
12/04/2014 12:05:57.
The Authors would like to thank Morten Johnsen Dille for general input, Ann-Sissel
Teialeret Ulset for performing the SEC-MALLS measurement and Ingvild Johanne Haug
for valuable discussions.
References
1. A. Veis, The macromolecular chemistry of gelatin, Academic Press, New York,, 1964.
2. W. Babel, Chem Unserer Zeit, 1996, 30, 86-95.
3. I. J. Haug, L. B. Sagmo, D. Zeiss, I. C. Olsen, K. I. Draget and T. Seternes, Eur J
Lipid Sci Tech, 2011, 113, 137-145.
4. F. Podczeck and B. E. Jones, Pharmaceutical capsules, 2nd edn., Pharmaceutical
Press, London, 2004.
5. J. Eysturskaro, I. J. Haug, A. S. Ulset and K. I. Draget, Food Hydrocolloid, 2009, 23,
2315-2321.
6. D. A. Ledward, in Functional properties of food macromolecules, eds. J. R. Mitchell
and D. A. Ledward, Elsevier Applied Science Publishers, London ; New York,
N.Y., 1985, pp. 171-201.
7. J. W. Janus, B. E. Tabor and R. L. R. Darlow, Kolloid-Z.u.Z.Polymere, 1965, 205,
134-139.
8. J. Eysturskard, I. J. Haug, N. Elharfaoui, M. Djabourov and K. I. Draget, Food
Hydrocolloid, 2009, 23, 1702-1711.
9. R. Schrieber and H. Gareis, Gelatine handbook : theory and industrial practice,
Wiley-VCH, John Wiley distributor, Weinheim Chichester, 2007.
DEVELOPMENT OF A DAIRY DESSERT WITH FUNCTIONAL PROPERTIES
1 INTRODUCTION
The current hectic lifestyle and poor diet adopted by many people is leading to an
increase in the occurrence of a number of medical conditions including diabetes, obesity,
atherosclerosis and pancreatitis. Sugar replacement and lowering the calorific value in
traditional products are key to the development of novel products for a healthy human diet.
It is well known that milk and dairy products play an important role in the human
diet due to their enhanced biological and nutritional value1. One of categories of dairy
products which are popular around the world are dairy desserts (creams, puddings,
cocktails, whipped cream), and the most important factor for this group of products is their
rheological properties (viscous, gel-like). Utilizing whey, a by-product of cheese
12/04/2014 12:05:59.
manufacture, in novel product development provides good nutrition to the population, and
leads to a decrease in ingredient costs and minimizes the expense for recycling. Skimmed
milk is also widely used as a basic ingredient in dairy products which also leads to a
reduction of fat content in finished products2.
In the dairy industry, manufacturing of dairy desserts requires the use of a wide
range of ingredients to control the consistency. Presently, utilizing such ingredients is
sufficient since they may act not only as technological agents, but also provide health
benefits. Polysaccharides find broad application in the food industry and in addition to
proteins, are the main components that define food structure and functionality. The
interaction of polysaccharides with water, proteins and other food ingredients has an
influence on both the rheological and organoleptic properties of food products 3-5.
Currently, there is extensive research being undertaken in the area of product
development that could provide delivery of essential nutrients, i.e. dietary fiber, vitamins,
minerals, fatty acids, antioxidants, and prevent various diseases. Among the broad
spectrum of ingredients with nutritional value, soy is of particular importance6. This
research deals with the introduction of dietary fiber (in the form of polysaccharides) and
soy protein in the development of dairy desserts utilizing whey.
The materials utilized in the study were: low methoxy pectin (LMP) (FMC, USA),
locust bean gum (LBG) (FMC, USA), L-carrageenan (FMC ,USA), xanthan gum (CP
Kelco ApS, Denmark), soy protein isolate (Shandong Jianyuan Foods, China), gelatin
(Veles, Russia), corn starch (Rospak, Russia), fructose (Dietary products, Russia), betulin
extract of birch bark (Birch world, Moscow).
View Online
2.2Methods
volume of the glass. The pH of the studied systems was determined at room temperature
using the "Checker" (Hanna, USA). Sensory evaluation of the puddings was performed by
15 panelists using a hedonic scale where panelists were asked on the level of preference9.
The nutritional assessment for the developed products was carried out using information
provided in the literature10.
Data from Table 1 show that all polysaccharides used in the study fully dissolve in
water giving solutions with different pH values. This information allows us further to
manipulate polysaccharide compositions and assess their potential behavior in the food
systems.
The viscosity of solutions of polysaccharides and their binary systems used as
thickening agents were studied and the results are reported in Figure 1.
Data from Figure 1 show that in order to obtain the identical viscosity to gelatin the
concentration of polysaccharides should be up to six times less.
There has been considerable interest in studying the interactions in binary
hydrocolloid mixtures and it is now recognized that in systems where the hydrocolloids
associate it can result in gel formation or precipitation. Hydrocolloids with opposite
charges most likely will associate with the creation of precipitates, while association of
some rigid polysaccharide molecules will lead to a gel formation11,12. In the case where
there is no association between hydrocolloids the system will exist as a single
homogeneous phase at low concentration, whereas at higher concentration the system will
separate into two liquid phases over time and each phase will be enriched with one of
hydrocolloids. Thus, the choice of the type and concentration of hydrocolloids will result
in the formation of various structures13 - 16.
Based on the obtained data, the binary compositions chosen for the production of
dairy desserts were as follows: 0.3% for -carrageenan, 0.3% for xanthan gum, 0.8% for
LBG, 0.4% for the mixture of LMP and -carrageenan (1:1), 0.15% for the mixture of
xanthan gum and LBG (0.5:1).
12/04/2014 12:05:59.
Figure 1: Absolute viscosity for -carrageenan (+), a mixture of -carrageenan with LMP
(1:1) (), a mixture of -carrageenan with LBG (1:0.5) (U), a mixture of -carrageenan
with xanthan gum (1:1) (), a mixture of xanthan with LBG (0.5:1) (), xanthan gum
({), LBG (), LMP (S), and gelatin (z) at 22 C.
View Online
3.2 Development novel formulation concepts for the dairy puddings with
functional properties
Puddings with
Ingredients L- LMP and L- xanthan and
gelatin LBG xanthan
carrageenan carrageenan LBG
Liquid whey 57.0 62.0 60.0 60.0 60.0 60.0
12/04/2014 12:05:59.
4 CONCLUSIONS
suitable for vegetarians due to the absence of gelatin and enriched with antioxidants. The
sensory evaluation test performed among fifteen panelists show the acceptability ofthe
products which is extremely encouraging for the future development work.
Novel puddings can be recommended for the people with specific nutritional needs
as well as for medical purposes since the product does not contain sugar and is enriched
with essential components.
5
Sensory evaluation, score
0
gelatin i-carrageenan xanthan plus i-carrageenan LBG xanthan
LBG (0.5:1) plus LMP
(1:1)
Figure 2: Sensory evaluation for the developed dairy puddings as compared to the
pudding with gelatin.
View Online
5. ACKNOWLEDGMENTS
uthors would like to thank Professor Peter A. Williams and Dr Graham Sworn for
samples of polysacharides.
References: