Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
PHASE BEHAVIOUR AND MODELING OF LOW AND HIGH-SOLID
BIOPOLYMER MIXTURES: A TREATISE
PREETI SHRINIVAS
(B.Tech, U.I.C.T.)
A THESIS SUBMITTED FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF CHEMISTRY
NATIONAL UNIVERSITY OF SINGAPORE
2009
ii
ACKNOWLEDGEMENTS_______________________________ __
Successful completion of this thesis would not have been possible without the
research scholarship offered to me by NUS through the Food Science and Technology
Programme. I therefore take this opportunity to express heartfelt gratitude and
appreciation to have had this privilege.
I am greatly indebted my guide and mentor, Prof. Stefan Kasapis whose constant
and immense guidance, support and encouragement has been pivotal. I consider myself
extremely fortunate to have had you as my supervisor and thank you for helping me
endure patiently and sail through a difficult yet worthwhile and memorable journey!
I am extremely grateful to my current supervisor Prof. Liu Shao Quan for taking me
on as his student and assisting me through the final year. I would also like to extend my
gratitude to the other faculty members and staff of the Food Science department for their
valuable inputs and suggestions. In particular, I am grateful to Ms. Lee Chooi Lan, Ms.
Huey Lee and Rahman.
I would also like to take this opportunity to thank Mr. William Lee and Mr. Derrick
for their technical assistance, Mr. Abel Gaspar Rosas for his constant encouragement and
the staff at the Dept. of Chemistry, Dept. of Biological Sciences and IMRE, for
permitting usage of their facilities.
I owe a big thank you to Limei, Cynthia and Denyse for whole heartedly
participating in this project, as also to my friends and fellow students at FST.
iii
I would next like to thank all those people whom I love immensely but haven’t
expressed it often enough-
Mom and Dad, the very reason I have reached this far in life!
Sujit, Sunil, Praveen, Rashmi and Archana- my anchors and pillars of strength.
Jatin- through your encouragement I began this journey and in many ways you are the
reason for its completion.
My grandparents, uncles, aunts, cousins and well wishers whom I have not mentioned by
name here.
Dr. Ramamoorthy, Dr. Rao, Archana S. for being there when I needed you the most.
My friends- old and new, Jiang Bin, Lilia, Shen Siung, Jorry, Mya, Neha, Sumantra,
Tanmay and all those who have made Singapore home for me.
Jayanth- for being extremely kind, understanding and supportive.
Dinesh- for blessing me and being with me.
Sadhguru- whom I deeply revere.
And finally, God- for showing me who He really is.
iv
TABLE OF CONTENTS__ _____________________________ Page
ACKNOWLEDGEMENTS ii
SUMMARY vii
LIST OF TABLES ix
LIST OF FIGURES x
LIST OF ABBREVIATIONS xiii
LIST OF PRESENTATIONS AND PUBLICATIONS xiv
PREFACE xv
PART I
MECHANICAL PROPERTIES AND PHASE MODEL INTERPRETATION
OF A COMPOSITE SYSTEM COMPRISING GELATIN,
AGAROSE AND A LIPID PHASE
CHAPTER 1: INTRODUCTION 2
1.1 Gelatin 4
1.2 Agarose 6
1.3 Biopolymer Mixtures 8
1.4 Phase Separation in Biopolymer Mixtures 9
1.5 Polymer Blending Laws of Takayanagi 11
1.6 Davies Law of Bicontinuity 15
v
CHAPTER 2: THEORY
2.1 Rheology 18
2.2 Differential Scanning Calorimetry 21
2.3 Scanning Electron Microscopy 22
CHAPTER 3: EXPERIMENTAL SECTION
3.1 Materials 23
3.2 Methods 25
CHAPTER 4: RESULTS AND DISCUSSION
4.1 Experimental Observations on Single Preparations of Gelatin, 28
Agarose and Lipids used
4.2 Experimental Observations on Mixed Systems of Gelatin, Agarose 36
and a Lipid Phase
4.3 Quantitative Analysis of Mechanical Functions in Support of the 48
Phase Topology of the Agarose/Gelatin/Lipid Mixture
CHAPTER 5: CONCLUSIONS 58
CHAPTER 6: SUGGESTIONS FOR FUTURE WORK 59
REFERENCES 60
vi
PART II
EFFECT OF CO-SOLUTE (GLUCOSE SYRUP) ON THE
STRUCTURAL BEHAVIOUR OF AMYLOSE GELS
CHAPTER 1: INTRODUCTION 68
1.1 Amylose 69
1.2 Glass Transition in High Solid Systems 71
CHAPTER 2: EXPERIMENTAL SECTION
2.1 Materials 78
2.2 Methods 79
CHAPTER 3: RESULTS AND DISCUSSION
3.1 Qualitative Aspects of the Effect of Increasing Levels of Co-Solute 82
on the Structural Properties of Amylose
3.2 Amylose Diverging from the Paradigm of Coil-to-Helix 86
Polysaccharides in a High Solids Environment
3.3 Utilization of the Method of Reduced Variables to Quantify the 94
Viscoelasticity of Amylose-Sugar Mixtures during Vitrification
CHAPTER 4: CONCLUSIONS & FUTURE TRENDS 103
REFERENCES 105
vii
SUMMARY:
The first part of this thesis attempts at examining the structural properties of binary
and tertiary mixtures made of the cold-setting biopolymers agarose and gelatin, and a
lipid phase with solid or liquid-like viscoelasticity. The working protocol included the
techniques of small-deformation dynamic oscillation on shear, modulated differential
scanning calorimetry and scanning electron microscopy, and theoretical modeling that
adapted ideas of relating morphology to elastic modulus of synthetic polyblends and
block polymers. The experimental setting was designed to encourage extensive phase
separation in the binary gel of agarose and gelatin whose mechanical properties were
rationalized on the basis of a bicontinuous blending-law. The presence of two continuous
phases allowed the slower-gelling component (gelatin) to exhibit favourable relative
affinity for solvent with increasing concentrations of the protein in the system. This is an
unexpected outcome that contradicts the central finding of a single value of the “p-factor”
observed in the distribution of solvent between the continuous matrix and discontinuous
inclusions of de-swelled binary gels reported earlier in the literature. Incorporation of a
lipid phase of effectively zero elastic modulus or in excess of 108 Pa in the composite
aqueous gel weakens or reinforces the matrix accordingly. The elastic moduli and
morphology of the tertiary blend were related to changing the relative phase volumes of
components using analytical expressions of isotropically dispersed soft or rigid filler
particles in a polymeric matrix.
The second half of the thesis presents data concerning the structural behavior of
amylose in the presence of glucose syrup and a possible interpretation of the same.
Observations were obtained once again by the aforementioned experimental methods of
viii
small-deformation dynamic oscillation on shear, modulated differential scanning
calorimetry and scanning electron microscopy. In contrast to industrial polysaccharides
that undergo readily a coil-to-helix transition (e.g., agarose, deacylated gellan and κ-
carrageenan), amylose holds its structural characteristics unaltered at low and
intermediate levels of glucose syrup. This is followed by an early phase inversion from
polysaccharide to co-solute dominated system at levels of solids above 70.0%, whereas
industrial polysaccharides can dictate kinetics of vitrification at levels of solids as high as
90.0% in the formulation. Additional viscoelastic “anomalies” include a clear breakdown
of thermorheological simplicity with data exhibiting two tan δ peaks in the passage from
the softening dispersion to the glassy state. Besides phenomenological evidence,
mechanistic modeling using the combined framework of the free volume / reaction rate
theories argue for two distinct glass transition temperatures in the mixture. It is proposed
that the amylose / glucose syrup / water system does not reach a state of molecular
mixing, with the morphological features being those of a micro phase-separated material.
ix
LIST OF TABLES
PART I
Table 3.1 Composition of Soybean Oil 24
Table 3.2 Composition of Hydrogentated Vegetable Fat 24
x
LIST OF FIGURES
PART I
Figure 1.1 Structure of agarose
6
Figure 1.2 Schematic representations of ideal rubber, gelatin and agarose
7
Figure 1.3 Changes in calculated modulus as a function of SX
15
Figure 4.1 Storage and loss modulus variation as a function of
temperature and time of observation for gelatin and agarose
on cooling to 0oC
29-30
Figure 4.2 Storage and loss modulus variation as a function of
temperature and time of observation for gelatin and agarose
on cooling to 25oC
31
Figure 4.3 Calibration curves of storage modulus as a function of
polymer concentration for agarose and gelatin at 0 and 25°C
33
Figure 4.4 Storage and loss modulus variation as a function of
temperature and time of observation for Dalda Vanaspati lipid
35
Figure 4.5 Complex viscosity variation as a function of shear rate for
soybean oil at 25oC
36
Figure 4.6 Heating profiles of storage and loss modulus for mixtures of
gelatin, agarose and a lipid
38
Figure 4.7 DSC exotherms for single, binary and tertiary mixtures
40
Figure 4.8 Scanning electron microscopy images of binary and tertiary
mixtures
42
Figure 4.9 Master curves of experimental storage modulus data obtained
at 0oC and 25°C for single and binary mixtures
44
Figure 4.10 Experimental storage modulus data obtained as a function of
lipid concentration in tertiary mixtures at 0oC and 25°C
47
Figure 4.11 a) Modeling the phase topology of the agarose/gelatin gel at
25°C using the isostrain and isostress blending laws for a
binary sample comprising 1% agarose plus 10% gelatin
51
xi
Figure 4.11 b) Modeling the phase topology of the agarose/gelatin gel at
25°C using the Davies blending law for bicontinuous
composite gels
53
Figure 4.11 c) Modeling the phase topology of the agarose/gelatin gel at
25°C using predictions of the “p factor” based on the
bicontinuous blending law for the composite gels of Figure
4.11b
55
PART II
Figure 3.1 Storage modulus variation as a function of time of observation
at 25°C for amylose gels of different concentrations
83
Figure 3.2 Frequency variation of storage modulus, loss modulus and
complex viscosity for 4.0% amylose plus 30.0% glucose
syrup and 2.0% amylose plus 70% glucose syrup at 25°C
85
Figure 3.3 Variation of normalized storage modulus on shear as a
function of sugar concentration coil-to-helix polysaccharides
and single sugar preparations
87
Figure 3.4 Heat flow variation as a function of temperature for amylose,
amylose/glucose syrup and glucose syrup samples
89
Figure 3.5 Cooling run of storage and loss modulus for 2.0% amylose in
the presence of 78.0% glucose syrup
92
Figure 3.6 Scanning electron micrographs of single amylose gels and in
combination with different concentrations of co-solute
93
Figure 3.7 Cooling run of storage modulus, loss modulus and their ratio
(tan δ) for 2.0% amylose in the presence of 70.0% glucose
syrup
95
Figure 3.8 Frequency variation of storage and loss modulus for 2.0%
amylose plus 70.0% glucose syrup at select temperatures
97
Figure 3.9 Master curve of reduced shear moduli (G'p and G"p) for the
sample of 2.0% amylose plus 70.0% glucose syrup as a
function of reduced frequency of oscillation (ωaT) based on
the frequency sweeps of Figure 3.8
98
xii
Figure 3.10 Temperature variation of the factor aT within the glass
transition region of amylose, glucose syrup and the glassy
state of 2.0% amylose plus 70.0% glucose syrup, reflecting
the WLF and modified Arrhenius fits of the shift factors
throughout the vitrification regime
101
xiii
LIST OF ABBREVIATIONS
G’ Storage Modulus
G’’ Loss Modulus
DSC Differential Scanning Calorimetry
MDSC Modulated Differential Scanning Calorimetry
SEM Scanning Electron Microscopy
ARES Advanced Rheometric Expansion System
LVR Linear Viscoelastic Region
Tg Glass Transition Temperature
TTS Time-Temperature Superposition
DE Dextrose Equivalent
DP Degree of Polymerization
KOH Potassium Hydroxide
GDL Glucono-δ-lactone
HCl Hydrochloric Acid
WLF William Landel Ferry
xiv
LIST OF PRESENTATIONS AND PUBLICATIONS
1. Shrinivas P., Chong L-M., Tongdang T. and Kasapis S. “Structural Properties and
Phase Model Interpretation of the Tertiary System Comprising Gelatin, Agarose and
Lipids. Part I: Inclusion of the Oil Phase”.
Poster presentation at the 8th
International Hydrocolloids Conference held in
Trondheim, Norway, (June ’06).
2. Shrinivas P., Tongdang T. and Kasapis S. “Structural Properties and Phase Model
Interpretation of the Tertiary System Comprising Gelatin, Agarose and a Lipid Phase”.
Oral presentation/proceedings submission at the 14th
Gums and Stabilizers for the
Food Industry Conference held in Wrexham, UK, (June ’07).
3. Shrinivas P., DeSilva D. and Kasapis S. “Effect of Co-solute (glucose syrup solids)
on the Structural Behaviour of Amylose Gels”.
Poster presentation at the 9th
International Hydrocolloids Conference held in
Singapore, (June ’08). Awarded Best Poster.
4. Shrinivas P., Kasapis S. and Tongdang T. (2009). “Morphology and Mechanical
Properties of Bicontinuous Gels of Agarose and Gelatin and the Effect of Added Lipid
Phase”. Langmuir, 25 (15), 8763-8773.
5. Shrinivas P. and Kasapis S. (2010). “Unexpected Phase Behaviour of Amylose in a
High Solids Environment”. Biomacromolecules, 11 (2), 421-429.
6. Kasapis S. and Shrinivas P. (2010). "Combined Use of Thermomechanics and UV
Spectroscopy to Rationalize the Kinetics of a Bioactive-Compound (Caffeine)
Mobility in a High Solids Matrix". Journal of Agricultural and Food Chemistry,
American Chemical Society, (in press- online access DOI: 10.1021/jf904073g).
7. Torley P.J., de Boer J., Kasapis S., Shrinivas P., Jiang B. (2008). “Application of the
Synthetic Polymer Approach to the Glass Transition of Fruit Leathers”. Journal of
Food Engineering, 86, 2, 243-250
xv
PREFACE:
The term ‘Biopolymers’ refers to a wide range of polymers of biological origin. It
encompasses all naturally available polymeric macromolecules such as proteins,
polysaccharides, lipids, nucleic acids. Each biopolymer is typically made up of a large
number of repetitive monomer units which could be sugars or amino acids. The chemical
composition and sequence in which these units are arranged are inherently well defined
giving rise to a basic ‘primary structure’. Some biopolymers, like proteins, fold into
characteristic shapes giving rise to secondary and tertiary structures. The molecular mass
distribution of a biopolymer depends on the type and the manner in which it is
synthesized. Accordingly, they may be classified as ‘monodisperse’ or ‘polydisperse’.
One or more macromolecular types are involved in most biological structures and
processes. Therefore, the presence/absence of chemical interactions between
macromolecules in a mixture and their resulting behaviour corroborates the sphere of
biological sciences. Numerous examples exist in the realm of applied sciences that
elucidate the use of macromolecular mixtures to produce favourable effects. Drug
delivery, food processing and technology are but a few emerging areas that involve
applications of such biopolymer synergism. Several food applications entail the use of
protein and polysaccharide mixtures to provide improved structure, mouthfeel,
processability and storage stability. The addition of even a small amount of a different
component (eg. sugar), can enhance the properties of proteins/polysaccharides
(gelling/thickening ability). Such interactions taking place in binary systems guide the
development of low fat spreads, confections and processed fish products. Designing
suitable macromolecular systems enables efficient drug delivery in pharmaceutical
xvi
sciences. Drug/capsule matrices made of biopolymer and co-solute interact favourably
with macromolecules (glycoproteins, plasma proteins). This is achieved by controlling
the mobility transition temperature of residual water (maintained below the glass
transition temperature Tg), resulting in a glassy matrix that specifically interacts with the
active compound.1,2
Thus, reports this far suggest that mixing two macromolecules results in one of the
following events taking place:
i. Absence of interactions/reactions of any kind.
ii. Phase separation due to thermodynamic incompatibility.
iii. Covalent/non-covalent interactions in a reversible/non-reversible manner.
However, more often than not, effects are observed on mixing macromolecules.
This in turn has generated a greater interest in cases involving such interactions than in
those that are devoid of any. Amongst the several biopolymers present in nature, the
focus of discussion in this thesis will be on three biopolymers in particular: Gelatin, a
protein; agarose and amylose, both of which are polysaccharides. In addition, the
properties of and roles played by sugars and lipids in composite biopolymer systems will
be discussed. The first part of this thesis will deal with a low solids biopolymer system
comprising primarily of gelatin, agarose and a lipid. An attempt is made thereof to
address issues stemming from the phenomena of phase separation, as also from filler
effects within biopolymer composites. The second part will focus on progression from a
low-solids to a high-solids environment wherein amylose will be the biopolymer of
xvii
interest. The effect that glucose syrup solids have (as a co-solute) on the structural
properties of amylose gels will be examined in detail.
Thus, an attempt is made through this thesis to address two distinct types of effects
observed in biopolymer mixtures, one wherein molecules ‘push apart’, while another
where they interact in a simple associative or irreversible aggregative manner to ‘stick
together’. This dividing line however, is not rigorous since several concepts used apply to
both parts.2
PART I
MECHANICAL PROPERTIES AND PHASE MODEL INTERPRETATION
OF A COMPOSITE SYSTEM COMPRISING GELATIN,
AGAROSE AND A LIPID PHASE
2
CHAPTER 1: INTRODUCTION
The phenomenon of gel formation by biopolymers such as proteins and
polysaccharides is widely known and has been a subject of interest to many academicians
and scientists in the last few decades. Structural manipulation of products using gelling
biopolymers to obtain varieties of textures and profiles is commonly practiced in the
food, beverage and pharmaceutical industries. As ever, the industrialist is faced with the
challenge of innovation in an increasingly competitive market in terms of ingredient cost,
product added-value, and expectations of a healthy lifestyle to mention but a few.3 An
outcome of this is the gaining popularity of the usage of proteins and polysaccharides as
stabilizers, thickeners or gelling agents in the production of commercial low fat spreads, a
preferred alternative to butter in recent times. Unlike butter, consumption of which has
been associated with an increased risk to heart disease and other related ailments, low fat
spreads mimic the texture and spreadability of butter and at the same time lower such
risk. To this effect, polysaccharides such as ‘Starch Hydrolysis Products’ mimic the
organoleptic properties of fat to a large extent. However, they provide a ‘starchy’
mouthfeel which is not desirable and hence addition of gelatin is invariably resorted to.
Gelatin is known to impart a melt in the mouth property to foods and thereby helps
improve the mouthfeel and flavour release characteristics in food systems.4 The
organoleptic properties of such a protein-polysaccharide mixture could potentially be
enhanced by further incorporation of a lipid with the resultant combination proving to be
ideal in meeting the varied requirements of low fat spreads.
3
Amongst all those known, thermal setting as a method of inducing gel formation
using a variety of gelling biopolymers has probably been the most widely investigated.5
Extensive work on mechanical characterization of gels has led to an understanding of the
mechanism of formation of gel structures as well as the resultant network properties. By
and large, this has enabled to distinguish between the gelation behaviour of
polysaccharides and globular proteins as cold setting and heat setting respectively.
Gelatin remains to be an exception however, as although it is a protein, it exhibits
gelation on cooling.
The food industry today has seen tremendous usage of several such biopolymers in
various combinations and proportions to obtain desired stability, performance and
consistency of almost every other food product being manufactured. A clear
understanding of interactions between biopolymers in the sol and gel states is therefore
necessary in order to control their behaviour and properties in multicomponent systems.
Temperature, salt content, charge, molecular weight, conformational ordering and
gelation are parameters that have a direct influence on the microstructure and rheology of
biopolymer composites as well as the likelihood of phase separation.6 Numerous mixed
systems of proteins and polysaccharides have been examined for their rheological
properties and applications to food products. Pioneering research to this effect was
carried out using a gelatin-agarose model composite. As much as it formed the basis to
advance investigations using other biopolymers in different combinations, loopholes for
the gelatin-agarose system remain to be questioned. Furthermore, numerous works on
protein and polysaccharide composite gels did not evolve into studies of systems
containing lipids, an outcome that would have been a natural development in our view.
4
This study therefore aims at investigating the effect of incorporation of a lipid phase on
the stability, gelling behavior and mechanical profile of a three phase system comprising
gelatin, agarose and a lipid. The present chapter will discuss in accordance, the properties
of the individual components of the chosen system, followed by a comprehensive
overview of concepts, theories and literature relevant to this study.
1.1 Gelatin:
A protein derived from animal sources, gelatin has proven to be one of the most
popular hydrocolloids till date with a varied range of applications in the food industry.
Commercially, it is derived from collagen, by controlled acid or alkaline hydrolysis
giving two distinct types of gelatin. The properties of gelatin thus depend on the source,
age and type of collagen used in its manufacture. Unlike polysaccharides, it melts at a
lower temperature. The unique ‘melt in the mouth’ property of gelatin can be attributed to
this factor.
Each collagen molecule is a triple helix made up of a 3 dimensional structure of α
chains with glycine occupying every third residue alternating with imino acids proline
and hydroxyproline which are known to impart rigidity to the molecule. The helix is
stabilized by interchain hydrogen bonding. Collagen can thus be characterized by the
following distinguishing features:
• High percentage of glycine (~33%)
• High proportion of imino acids hydroxyproline and proline (~22%)
• Repeating triplets of the gly-X-Y sequence where a high proportion of X and Y
are proline and hydroxyproline.
5
Gelatin closely resembles the parent collagen in primary structure. However,
pretreatment and extraction procedures give rise to minor differences such as:
• An increase in the aspartic and glutamic content lowering the isoelectric point
primarily due to increase in the number of carboxyl groups.
• Conversion of arginine to ornithine in prolonged treatments
• Lower proportion of trace amino acids such as cysteine, tyrosine than in parent
collagen.
At low temperatures, a conformational disorder-order transition is believed to
occur, yielding thermoreversible networks created by triple helix formation of gelatin
chains with distinct cross-linked junction zones, stabilized by hydrogen bonding.
Formations of ordered quasi-crystalline triple helical junction zones separated along a
single polymer chain contour by flexible regions characterize the initial gelation of
gelatin.7 The thermal stability of gelatin depends on the concentration used. On lowering
the temperature, the helix content increases, however, the helix stability is reduced. The
energy needed to melt a gelatin gel is related to the number of junction zones and their
thermal stability.8
In the food industry, gelatin is used for a great number of applications such as
confectionery and desserts, dairy products, meat products, hydrolyzed gelatin
applications, sauces, dressings, wine fining and many more. Gelatin is regarded as a
multipurpose ingredient as it can be used as a gelling agent, whipping agent, stabilizer,
emulsifier, thickener, adhesive, binder or fining agent.
6
1.2 Agarose:
Fig. 1.1 Structure of Agarose (Source: Ref. 5)
Agarose, the gelling component of Agar, is a neutral polysaccharide obtained from
a family of red seaweeds (Rhodophyceae). As opposed to agaropectin, the charged
polysaccharide fraction of agar, the sulphur content of agarose is negligible. Agarose has
a linear structure with no branching, which infact is quite similar to structures of kappa
and iota carrageenans except for the sulfate content and L configuration. In the solid state
it exists as a threefold, left handed double-helix. A central cavity lined with hydroxyl
groups exists along the helix axis enabling hydrogen bonding with water. It is not soluble
in cold water but dissolves completely in boiling water. Prior soaking helps reduce
dissolution time. Such hydration unleashes fluctuating, disordered coils. The maximum
concentration that can be obtained by normal dissolution in water is approximately 3-4%
as it a high molecular weight material. It has the ability to form gels at very low
concentrations. A disordered, random coil structure at elevated temperatures cools to
form a gel network at low polymer concentrations by adopting an ordered double-helix
state. Occurrence of ‘kinking’ residues terminates helix formation, causing interchaining
of different agarose molecules to give a three-dimensional network.9 The structural knots
of such an enthalpic network are highly aggregated intermolecular associations composed
of a plethora of coaxial double helices as postulated on the basis of x-ray fiber diffraction
and optical rotation.
7
The gelling phenomenon is completely reversible. Dissolution of the polymer at
temperatures as high as 90oC followed by cooling induces gel formation below 40oC
depending on the concentration of the polymer used. Gels thus formed melt on reheating
at temperatures just below the boiling point of water. Heating however unveils substantial
thermal hysterisis, i.e. a difference/lag between gelling and melting temperatures, another
important feature of agarose gels. The gel hysteresis of agar greatly exceeds that of other
gelling agents and is the basis for many of its applications in food and biotechnology.
In the food industry agarose is useful in applications such as low calorie foods. It is
also used as a gelling agent. Agarose gels are also used for the gel electrophoresis
technique used in biotechnology.
In their work in 1985,28 McEvoy and his research group showed how networks of
Gelatin and Agarose differed from that of ideal rubber as seen in Fig. 1.2
Fig. 1.2 Schematic representations of network types: a) Ideal rubber. Circles are covalent
crosslinks. b) Gelatin and c) Agarose. Heavy lines indicate helical junction zones.
(Source: Ref. 28 with permission)
8
1.3 Biopolymer Mixtures
Whilst studies on single biopolymer systems have enabled their characterization,
extensive work has been carried out on biopolymer mixtures as well. One of the
fundamental assumptions here being- a biopolymer mixture is composed of at least two
different biopolymers in addition to an aqueous component which forms the largest
component in terms of volume. The type of resultant mixture depends on the nature of the
biopolymers. Various possibilities arise, such as, a fluid-fluid system, a solid dispersed in
a fluid or a mixed gel system in which gelation of both biopolymers has occurred.10
Depending on the nature of the network developed, the third case can further be classified
as
1. Interpenetrating networks: Here, each of the biopolymer gels separately
forming independent networks that interpenetrate into one another.
2. Coupled networks: In this case, the polymers directly associate to form a single
network which could be characterized by covalent linkages, ionic interactions or
co-operative junction zones.
3. Phase separated networks: This is the third case which is observed more often
than not in mixed biopolymer mixtures. Above a certain critical concentration,
solutions of two different polymers usually phase separate due to thermodynamic
incompatibility that results in less favourable interactions between different
polymer segments. This causes each biopolymer to exclude the other from its
polymeric domain, thereby raising the effective concentrations of both. Phase
separation in protein-polysaccharide-water systems is usually known to occur
only when the total polymer concentration exceeds 4%. If the concentration of
9
one polymer is held constant, phase inversion occurs when the concentration of
the other exceeds a certain minimum value. The nature of the continuous
supporting phase is determined by examining the gel microstructure before and
after phase inversion.11
Formation of a binary gel is possible from both single-phase and phase separated
solutions leading to associative or segregative systems. In the former, gelation occurs
through electrostatic interactions between polyanion polysaccharides and polycation
proteins, or heterotypic junction formation between conformationally compatible
sequences of two polysaccharides.12-13
The latter is far more common and in the case of
the gelatin/agarose mixture the tendency of individual molecules to be surrounded by
others of the same type leads to composite gels where phase inversion, with increasing
concentrations of a given component (gelatin in this case), occurs at a specific mixture
composition.5
1.4 Phase Separation in biopolymer mixtures
The phenomenon of phase separation as mentioned earlier is a common feature in
biopolymer mixtures. It remains to be one of the basic tools of achieving the required
structural properties in a variety of industrial products. In terms of mechanistic
understanding of phase behaviour, an early advance was the appreciation that classic
phase-separation in solution leading to thermodynamic equilibrium between two polymer
phases is not carried over to the gel state.14
Formation of such a resultant composite gel is
accomplished by disorder-to-order transitions and possible aggregation of polymeric
segments of the two constituents on cooling of the polymeric solution to low
10
temperatures. The difference in free energy between the single phase and biphasic states
has been attributed to the enthalpy-entropy balance. Unlike for solutions of small
molecules, the enthalpy of interaction outweighs a grossly reduced entropic advantage of
mixing in the case of polymer solutions. Favourable interactions between two
biopolymers could result in a single gel phase. However, more often than not, interactions
between segments of two different polymers tend to be less favourable than those
between chains of any single type.15-18
This results in thermodynamic incompatibility
between the two giving rise to phase separation. Thus, at low concentrations, the two
polymeric species may co-exist within a single aqueous phase. At higher concentrations
however, the result is a spontaneous separation into two discrete phases, each containing
majority of one polymer and little of the other. In systems where both polymers can form
a cross-linked network, the end result is a biphasic, composite gel.19
Mixing
conformationally dissimilar macromolecules produces such biphasic composite gels
whose water partition between the two phases is profoundly affected by the phase
inversion in the system. For gels under kinetic control- formation of a continuous
network (caused by a faster gelling component or due to excessive amounts of one
polymer) ‘freezes’ the system and prevents diffusion of water between the two phases.20
The antagonistic effect operating between ordering-aggregation leading to gelation
that arrests phase separation and the thermodynamic drive to extensive phase separation
can be manipulated by the cooling regime. Slow-cooled materials exhibit early phase
separation and gel reinforcement in the way observed only beyond the phase inversion
point for high concentrations of the quenched counterparts. The resulting composite gel is
under kinetic control and the nature of phase topology is determined by the thermal
11
treatment, which to a great extent negates the state of thermodynamic equilibrium of the
mixture in solution.21 Kinetically trapped binary gels can also be manipulated by utilizing
polymers of increasing molecular-weight distribution that can reduce dramatically the
amount of polysaccharide required for phase inversion in the presence of protein.22
Further structuring of the discontinuous filler is achieved by applying a shear flow to a
phase-separated biopolymer solution and then entrapping the anisotropic macrostructures
by gelling either or both phases. Depending on shear history, inclusions vary form
spheroids to elongated, or irregularly shaped particles formed at high shear stresses, an
outcome that may find application in the diffusional mobility of bioactive compounds
within a polymeric matrix.23
1.5 Polymer Blending Laws of Takayanagi11,15-19
A combination of two simple viscoelastic models (parallel and series) in proportion
to their phase volumes helps prediction of overall viscoelastic properties. Takayanagi, in
his work, verified such a model. Dynamic extensional measurements were performed on
samples composed of two layers of different synthetic polymers. The strain was imposed
either parallel or perpendicular to their sides.
This analysis is based on binary composites of pure, mutually insoluble, synthetic
polymers whose individual rheological properties are independent of the macroscopic
amounts present, and hence is an approximation. Thus, in a two component composite, if
X and Y are the two components having shear moduli GX and GY respectively,
mechanical properties of the composite may be derived from such individual systems
present at phase volume fractions φX and φY, where φX + φY = 1. Assuming extreme cases
12
of strain and stress distribution, two equations are derived providing upper and lower
bound limits for the value of shear modulus GC of a composite formed from X and Y.
If a weak material X is dispersed as discrete particles within a continuous matrix of
a stronger material Y, the overall shear modulus of the composite is related to the
corresponding moduli of the component phases by
GC = φXGX + φYGY ……(1.1)
Equation (1.1) applies to isostrain conditions wherein the strain is approximately
uniform throughout the material. It provides upper bound limits as the overall strength of
the composite is higher when the stronger component forms the continuous network. The
deformation of the weak filler is dictated by the response of the surrounding stronger
matrix. Hence, both components are deformed to the same extent.
Conversely, if a strong material is dispersed discontinuously within a continuous
matrix of a weaker material, the overall shear compliance of the composite is obtained as
the corresponding weighted average of the individual compliances
JC = JXφX + JYφY ……(1.2)
i.e. 1/GC = φX/GX + φY/GY ……(1.3)
Equation (1.3) refers to isostress conditions where the stress maybe regarded as
constant in both phases, describing a lower limit appropriate to a weaker continuous
phase, thereby providing lower bound limits. The strong filler is deformed less than the
surrounding matrix, and the stress acting upon it is limited to the resistance of the matrix
to the imposed deformation.
13
The effect of shear on agarose-gelatin gels has also been discussed by Brown et
al.55 Shearing (by varying the rotor speed) has been shown to bring about phase inversion
in the gelatin-agar mixture. Another noteworthy finding from this piece of research was
that increasing the shear results in smaller and more spherical particles.
In pure polymer composites phase volumes are determined directly by the amount
of each polymer present. In the case of water based biopolymer composites however,
presence of water as a third component introduces an additional complication. The phase
volumes will now depend on the way this solvent partitions itself between the two
biopolymer constituents. This in turn depends on the relative powers of the two
biopolymers to attract solvent which needs to be calculated before using Takayanagi’s
equations. Thus, in aqueous biopolymer gels, in addition to the amount of polymer
present, the solvent avidity or ‘water holding capacity’ of the polymers helps in
determining the phase volumes.
This problem was addressed by Clark and his group by introducing a relative
affinity parameter ‘p’ where
α/x ……(1.4)
(1-α)/y
α is the fraction of water associated with the X phase and (1-α) is the fraction associated
with the Y phase. x and y are masses of the two polymers. In other words, ‘p’ is the ratio
of solvent to polymer in one phase divided by the corresponding ratio in the other phase,
characterizing solvent partition. p<1 implies that Y is more solvent attracting than X,
while the opposite is true for p>1.5
Concepts of nominal and effective (on gelation and subsequent phase separation)
concentrations proposed by Clark provide more of an equilibrium treatment approach. By
14
allocating different values to ‘p’, the effective local polymer concentration in each phase
is estimated. Using suitable fits for the relationship between gel modulus and
concentration, the gel modulus of each phase is calculated from these adjusted
concentrations. Finally, the overall modulus is determined using the Takayanagi
approach.5
In his paper in 1992 however, Morris questioned such an approach due to
contradictory evidence from experimental studies of model systems. Unlike Clark, he
proposed that the final phase volumes may be under kinetic rather than equilibrium
control. This is primarily due to potential ‘freezing’ of the system by formation of one or
both of the polymer networks, largely eliminating re-partition of the solvent between the
two phases. Thus, in systems where one component is induced to gel before another by
appropriate temperature control, the overall modulus might actually be closer to that
anticipated from the nominal concentration rather than higher values predicted for the
biphasic composite.19
Subsequently, Kasapis and his group introduced a correction factor to this approach,
since in many systems the polymers constitute about a third of the total sample. Thus the
approximation of equating phase volume to the volume of water in each phase no longer
holds and the direct contribution of polymers to phase volume cannot be neglected. In
order to obtain the true phase volumes, relative weights were then adjusted for density
difference between the phases.18
Morris has described in his paper on solvent partition,19
a generalized approach to
calculating solvent distribution between the two biopolymer phases wherein changes in
calculated modulus are shown as a function of SX, the solvent fraction in the X phase. At
15
low values of SX, most of the water is present in the Y phase making polymer X
extremely concentrated and hence GX >> GY. Conversely, at high values of SX, GX <<
GY. At a particular value of SX however, the moduli of the two phases cross over to give
GX = GY = GU = GL. Upto this critical point, the ‘upper bound’ (GU) value would
correspond to a system with polymer X as the continuous matrix while the ‘lower bound’
(GL) value will correspond to a polymer Y continuous one. Beyond this critical point, at
higher values of SX, GU will relate to a polymer Y continuous matrix while GL to a
polymer X continuous one.11,19
Fig. 1.3 illustrates this explanation.
Fig. 1.3 Changes in calculated modulus as a function of SX (Source: Ref. 11 with
permission)
1.6 Davies law of bicontinuity
The Takayanagi model described earlier is a reliable way of calculating the overall
modulus of biphasic biopolymer systems in which one phase is continuous and the other
16
is dispersed, using the isostrain model when the stronger component forms the
continuous phase and the isostress model applied to the converse situation.
In some systems however, both phases remain continuous, either for a specific
concentration range of the two polymers used or for a particular combination of
biopolymers. Calculation of the overall modulus from the individual moduli of the
constituent phases is then based on the theoretical relationship proposed by Davies in
1971.24
The analytical expression used in such cases is shown in equation (1.5).25
(GC)1/5
= φX(GX)1/5
+ φY(GY)1/5
……(1.5)
Like the Takayanagi blending laws, this relationship was initially intended for
composites of condensed materials. Details for the derivation of equation (1.5) can be
found in literature which points out that it was first applied to the mechanical and
dielectric properties of condensed synthetic composites with a macroscopically
homogeneous consistency. Piculell, Nilsson and Muhrbeck were the first to apply it to
hydrated biogels of iota and kappa carrageenan in 1992. 26,27
Whichever the equation in use, modeling of biphasic gels is based on one of the
following two assumptions:
Bulk phase separation occurs (in solution) prior to gelation which subsequently takes
place independently in each phase. In such a case, the two polymers are confined entirely
to their respective phases. This sequence of phase separation followed by gelation was
first proposed in 1983 where an attempt to model a system comprising agarose and
gelatin was made using ideas of bulk phase separation and solvent partition.
17
Alternatively, McEvoy in 1985 used the classic deswelling theory to suggest that
agar solutions gel at a higher temperature than gelatin, forming a network across the
whole system at the nominal concentration of modulus Gnom.28 It is then taken to a higher
concentration (ceff
) of modulus Geff
by removal of water as gelatin gels. In other words, if
a gel network with essentially permanent crosslinks is formed at an initial concentration
ci, and is then taken to a lower or higher final concentration cf, by introduction or removal
of solvent (swelling or deswelling), then, from the classic swelling theory, the associated
change in modulus is given by:
Gi/Gf = (ci/cf)q
……(1.6)
q is derived from the classical Flory type swelling theory, and varies depending on the
polymeric network that undergoes swelling/deswelling.29,30
The phase behaviour and topology of gelatin/agarose gels is revisited presently in
order to examine the suitability of the classical blending laws or the theory of polymer
deswelling in relation to the patterns of solvent partition between the two components
and at two distinct temperatures of gelation. Furthermore, increasing understanding of the
molecular dynamics of proteins and polysaccharides leads to the desire to control the
functionality of complex biomaterials where chemical reaction pathways and enzymatic
processes are critical considerations. Examples for the same include the diffusion-
controlled substrate/enzyme interaction, the prevention of flavor/color degradation and
the stability of a lipid phase within the polymeric matrix.31-33
Thus, the second objective
of this work is to model the phase behaviour of tertiary mixtures containing a lipid phase
by triggering solid or liquid-like viscoelasticity with changing experimental temperature.
18
CHAPTER 2: THEORY
2.1. Rheology:
Small deformation studies using dynamic oscillation experiments on a rheometer is
one of the most popular techniques today and is widely used in studying the flow
behaviour and gelation properties of materials. Using this technique, measurements of the
storage modulus, viscous modulus, complex viscosity and wide number of other
parameters can be carried out. For viscoelastic systems such as gels, the storage and
viscous moduli are parameters of prime importance that help characterizing the system in
question. The storage modulus describes the storage of energy in a structure. In other
words, it defines how ‘solid-like’ a material is. Its magnitude depends on the number of
interactions between ingredients in the sample. The higher the number of such
interactions and the stronger the interactions, the higher is the value of G’. In contrast, the
viscous or ‘loss’ modulus, describes the part of energy lost as viscous dissipation. It is
related only to the number of interactions and virtually independent of their strength. The
greater the number of interactions in which friction can be created the larger is G”.
Rheological characterization is typically carried out by measuring the storage (G’)
and loss (G”) moduli as a function of temperature, time, frequency and strain using
parallel plate geometry. Accordingly, experiments are termed as ‘Temperature Ramps’,
‘Time Sweeps’, ‘Frequency Sweeps’ and ‘Strain Sweeps’.
19
Temperature Ramp
In cold setting gels, a cross over of G’ and G” takes place at a particular
temperature when the polymer solution is cooled at a predefined rate. At this critical
point G’ overtakes G”, marking the onset of sol-gel transition commonly known as
gelation. The ramp rate greatly influences the gelation profiles obtained through such a
temperature sweep experiment. Such a temperature sweep is carried out at a fixed
frequency and amplitude of oscillation.
Time Sweep
A temperature ramp is followed by a time sweep wherein the values of both moduli
increase as a function of time at a constant temperature for a certain period until
equilibrium is more or less achieved. However, the extent of increase in the values of G’
is much higher than that of G”, which infact is almost negligible. In the case of gels, a
curing time is usually given during which the storage modulus keeps developing until a
state of equilibrium is reached. The time taken to reach this state varies from gel to gel. In
agarose gels for example, like most other biopolymer systems, the coil to helix
transformation occurs very fast resembling a true first order phase transition. An
extremely short curing time is therefore almost always sufficient in the case of agarose
gels. For gelatin however, an initial phase lasting several hours is followed by a much
slower process that continues for a very long time. Thus, depending on the concentration,
gelatin gels are subjected to much longer curing times before a pseudo-equilibrium state
is achieved as an absolute equilibrium state is very difficult to access due to its dynamic
nature. The extreme importance of a time sweep is thus exhibited in that it determines if
20
the properties of a system are changing over the time of testing. Such a time sweep is
carried out by measuring material response at a fixed temperature, frequency and
amplitude of oscillation.
Frequency Sweep
A frequency sweep usually follows a time sweep and monitors the material
response to increasing frequency (rate of deformation) at a constant amplitude (strain or
stress) and temperature. Frequency is the time required to complete one oscillation. The
data obtained from frequency sweeps helps determine to a large extent the category under
which a given sample can be classified viz. a dilute solution, an entangled solution, a
weak gel or a strong gel. Derived parameters such as complex viscosity (η*) and tan δ
provide extremely useful information about the nature of the system being tested. In
addition, data from frequency sweeps is used in time-temperature superpositioning in
order to gauge long term properties or extremely high/low frequencies beyond the scope
of the instrument or reasonable experimental time. This concept uses a direct equivalency
between time (frequency of measurement) and temperature and will be discussed in detail
later.
Strain Sweep
A strain sweep helps determine the extent to which a sample undergoes
deformation. For all dynamic oscillation test measurements, in order to verify that the
results are ‘real’ and not merely artifacts, it is extremely vital that all the test are carried
out at an amplitude within the linear viscoelastic region of the sample. The principle
21
behind this is that if the deformation is small or applied sufficiently slowly, the molecular
arrangements are never far from equilibrium. The mechanical response is then just a
reflection of dynamic processes at the molecular level which go on constantly, even for a
system at equilibrium. Within this domain of linear viscoelasticity, the magnitudes of
stress and strain are related linearly, and the behaviour for any liquid is completely
described by a single function of time. Thus, the material response to an increasing
amplitude at a constant frequency and temperature is monitored during a strain sweep.
This is done to determine the LVR as all other tests are required to be carried out at an
amplitude found in the LVR. Samples are assumed to be stable before carrying out a
strain sweep. An unstable sample is subjected to a time sweep prior to the strain sweep to
determine stability. It must be noted however, that the linear region changes as a function
of frequency and temperature.
2.2. Differential Scanning Calorimetry
Differential Scanning Calorimetry (DSC) is one of the most popular and widely
used thermal analysis techniques. It measures the heat flow into or out of the sample.
Endothermic and exothermic transitions are measured as a function of time. Glass
transition/vitrification, melting, crystallization, heat capacity, thermoset curing, enthalpy
recovery are some of the transitions measured. Phenomena such as crosslinking and
protein denaturation are often studied using this technique.
The principle behind this technique is based on the temperature difference between
a sample and a reference pan monitored by a sample and reference thermocouple. On
heating or cooling a sample, a transition occurs resulting in a small temperature
22
difference between the sample and reference. This temperature difference is then
converted into heat flow, which is a measure of the amount of heat flowing in or out of
the sample.
2.3. Scanning Electron Microscopy
Electron microscopy uses an electron beam as against light which is used in optical
microscopy. It differs from light microscopy in that it has electromagnetic lenses and the
electron beam has to travel through vacuum. This helps in achieving much higher
magnifications of images than light microscopy permits. SEM is based on the following
principle. When an electron beam in the microscope hits the sample surface, different
kinds of electrons are given off due to interaction between the beam and sample.
Different kinds of detectors corresponding to the different kinds of electrons produced are
present enabling their detection. The secondary detector primarily helps in producing a
three dimensional secondary or scanning image. Others help in adding to the scanning
image certain detailed minute observations.
23
CHAPTER 3: EXPERIMENTAL SECTION
3.1 Materials
The gelatin sample was prepared especially for research from Sanofi Bio-
Industries, Baupte, Carentan, France. It was the first extract from a single batch of
pigskin produced by acidic hydrolysis of collagen (type A). Analytical characteristics of
the sample, which were determined by the manufacturer, have been published in full
before and, in brief, gel permeation chromatography was used to identify the percent
weight of ten molecular mass classes and the number average molecular weight (Mn ~ 68
kDa) of the sample.34
Furthermore, the Bloom value and isoelectric point (pI) were found
to be 305 and 8.7, respectively. The average moisture content was estimated to be 9.51 %
using a Mettler Toledo LJ16 moisture analyzer.
Agarose (Type 1-B) was purchased from Sigma-Aldrich, Gillingham, UK. It is a
material of high gel strength (G' ~ 34 kPa at 0ºC for the 2% polymer concentration).
Using aqueous size exclusion chromatography, the supplier determined the number
average molecular weight of this agarose sample (Mn ~ 120 kDa). Furthermore, moisture,
ash and sulfate contents were of less than 10%, 0.25% and 0.12%, respectively.
Soybean oil, procured from Sigma-Aldrich, Singapore, was used as the liquid filler
in the tertiary mixtures, while Vanaspati (commercially available as Dalda), a saturated
fat of Indian origin, obtained by chemical hydrogenation of vegetable oils, was used as
the solid filler. The latter is a material of high hardness (G' ~ 2 x108
Pa at 0ºC). Soy
lecithin procured locally from Suntop Enterprise, Singapore, was used as an emulsifier to
ensure a macroscopically homogenous blend, thus preventing bulk phase separation. The
24
fatty acid composition of the lipid phase is shown in Tables 3.1 and 3.2. Samples were
analyzed in triplicate and standard deviation values are accordingly reported.
Table 3.1. Composition of
Soy Bean Oil
Composition
Content
Avg. (%) Std. Dev.
C16:0 10.719 0.453
C18:0 4.097 0.298
C18.1n9c 21.400 0.227
C18:1n9t 1.646 0.032
C18:2n6c 54.416 1.098
C18:3n6 1.027 0.067
C20:1 1.010 0.058
C18:3n3 5.685 0.048
Total 100.000
Table 3.2. Composition of
Hydrogenated Vegetable Fat
Composition
Content
Avg.(%) Std. Dev.
C12:0 0.251 0.041
C14:0 0.134 0.007
C16:0 17.659 0.348
C18:0 13.434 0.269
C18:1n9t 53.858 0.746
C18:1n9c 9.698 0.493
C18:2n6t 1.013 0.025
C18:2n6c 2.468 0.538
C20:0 0.512 0.043
C20:1 0.126 0.012
C20:2 0.506 0.041
C22:0 0.471 0.095
Total 100.000
25
3.2 Methods
Gel Preparation
Pure agarose and gelatin, were dissolved in distilled water at 90oC and 55oC
respectively till clear solutions were obtained. Solutions of varying concentrations were
made to obtain individual calibration curves. Binary mixtures were made by maintaining
agarose at 1% and varying the gelatin concentration from 0-30%. For this, agarose was
first dissolved at 90oC and then the solution was cooled down to about 55
oC before
adding in gelatin. After addition and complete dissolution of gelatin, the desired final
composition was achieved by adjusting the proportion of water. The stirring speed of the
magnetic stirrer was maintained at approximately 300 rpm. Tertiary mixtures were made
by maintaining the concentration of agarose at 1%, gelatin at 10%, and varying the lipid
concentration from 0-30%. Binary mixtures were first prepared as described above.
However, before adding in the lipid phase, the temperature of the binary mixture was
further brought down to around 45oC and maintained thereafter. Appropriate proportions
of the lipid and about 0.2% lecithin, an emulsifier, were added into the mixture at this
temperature. The use of lecithin in such minute quantities prevented bulk phase
separation and ensured homogenous incorporation of the lipid phase. The stirring speed
was gradually increased from 300 rpm to about 1000 rpm in order to obtain a uniform,
smooth creamy tertiary mixture. In addition, for the tertiary mixtures, it was ensured that
the biopolymer matrix into which the lipid phase was incorporated remained identical in
terms of composition to that of the binary mixtures.
26
Measurement Techniques
Small and large deformation measurements were carried out on a controlled strain
rheometer (ARES- TA Instruments) using parallel plate geometry (40 mm diameter; 2
mm gap). Biopolymer solutions of desired compositions were prepared, loaded at 45oC
and allowed to gel in situ on the rheometer. Measurements of storage (G’) and loss (G”)
moduli as a function of temperature, time, frequency and strain were made.
Initial gelation was allowed to occur by cooling the solutions at a ramp rate of
1oC/min and frequency of 1 rad/sec to a final temperature of 25
oC or 0
oC in order to
observe effects of the soft or hard filler respectively. The level of strain maintained was
0.1% for agarose and 1.5% for gelatin. Subsequently, the development of moduli as a
function of time was monitored. When equilibrium was attained, i.e. the experimental
moduli had not increased more than 0.5% in a 15 min time interval; the samples were
subjected to an increasing frequency of oscillation (0.1-100 rad/sec) at a fixed amplitude
(0.1% or 1.5%). Alternatively, in order to determine the linear viscoelastic region (LVR)
of the samples, strain sweeps were conducted by increasing the amplitude of oscillation at
a fixed frequency (1 rad/sec). It was ensured that all measurements were carried out at
levels of strain well within the LVR. In order to maintain uniformity, time sweeps for 1
hour were conducted for all samples. A temperature ramp at 1oC/ min to the gel melting
temperature followed the frequency sweep, again at a fixed amplitude and frequency.
For Differential Scanning Calorimetry (DSC) and Scanning Electron Microscopy
(SEM), gels were prepared as described earlier and gelation was allowed to occur by
cooling the solutions to room temperature. Gel pieces were then used in the following
protocols.
27
DSC measurements were performed on MDSC Q100 (TA Instruments V9.4 Build
287). The instrument used a refrigerated cooling system to achieve temperatures down to
0oC and a nitrogen DSC purge cell at 25mL/min. Samples were loaded into and sealed in
hermetic aluminum pans. The DSC heat flow was calibrated using a traceable indium
standard and the heat capacity response using a sapphire standard. Initially, about 10 mg
of agarose, gelatin, lipid or mixtures of the same were heated to 95oC, 70
oC, 50
oC and
95oC respectively to eliminate erroneous phenomena due to thermal history during
sample preparation and loading. Samples were then cooled to 0oC and reheated to 90
oC at
1oC/min. Samples were analyzed at + 0.53
oC temperature amplitude of modulation and a
40 s period of modulation. The reference was an empty hermetically sealed aluminum
DSC pan.1 Measurements were carried out in triplicate yielding effectively overlapping
traces.
For SEM, gel cubes of about 10mm3 in size were freeze-dried for 4 days at -30
oC.
They were then glued onto copper stubs using carbon paint, a highly conducting medium.
Samples were then allowed to dry prior to being subjected to gold coating. Sputtering
was performed with a JFC-1100 gold sputterer at 0.2 Torr, 10mA and 1.2 kV, resulting in
the formation of a 20 nm thick gold coating over the exposed samples surface. Scanning
of these gold plated samples was then done using a scanning electron microscope (JEOL,
Model JSM-220A, Japan) at an accelerating voltage of 15 kV.
Error bars are not shown for the rheological work; however three replicates were
analyzed for each experimental preparation, results being readily reproducible with a 5%
error margin as a function of temperature or timescale of measurement.
28
CHAPTER 4: RESULTS & DISCUSSION
4.1 Experimental Observations on Single Preparations of Gelatin, Agarose and
Lipids used
Gelatin and agarose have a long history of industrial use in food, pharmaceutical,
biomedical, cosmetic and photographic applications that has triggered an extensive
program of fundamental research on their gelation and structural properties.35-37
Nevertheless, meaningful experimentation on the phase morphology of their binary gels
or in mixture with a lipid phase necessitates a certain characterization of the gelation
behaviour of the individual components. In this part of the work, dynamic oscillation on
shear was utilized to obtain a good volume of data on concentration-temperature
combinations for the two polymers in a reasonable time.
Figure 4.1a shows the asymmetric cooling and heating traces separated by the
isothermal run at 0°C for 20% gelatin scanned at a rate of 1°C/min. The usual frequency-
dependent markers of network formation (i.e., sharp rise in the G' trace) and collapse (G'
falling below G" during heating) can be used to pinpoint the formation and melting
temperatures of the gelatin gel, which appear to be 28.9 and 36.3°C, respectively. These
values are used more for operational convenience in rationalizing aspects of phase
continuity in the tertiary mixtures, rather than determining gelling point based on the
frequency-independent loss tangent. The thermal hysteresis in gelatin remains constant
between six and seven degrees throughout the examined concentration range, and this
pattern of behaviour argues that helices, as opposed to aggregation, are the main drive of
structure formation in gelatin. Detailed discussions on the gelation process of the protein
29
that is initiated by bringing together two strands of the same molecule (‘hairpin’
structure) with a third strand can be found in the literature, a brief description of which
has been made in the preceding sections.38,39
0
1
2
3
4
5
6
0 20 40 60 80 100 120 140 160
Time (min)
Lo
g (
mo
du
lus/
Pa)
-10
0
10
20
30
40
50
60
70
Tem
per
atu
re (
°C)
G'
G"
a
Fig. 4.1a. Storage (�) and loss (�) modulus variation as a function of temperature ( )
and time of observation for 20% gelatin. (scan rate: 1°C/min; frequency: 1 rad/s;
strain:1.5%; equilibration temperature: 0oC).
The general form of modulus development for gelatin is contrasted with the
structure formation of a 2% agarose preparation in Figure 4.1b. Fluctuating, disordered
coils of the material in an aqueous solution give rise upon cooling to high values of
storage modulus, which signify the transformation to a three-dimensional, ordered
30
structure;40
conformational transition commences at about 34.2°C. Subsequent heating
unveils substantial thermal hysteresis, with the network melting at a temperature close to
the boiling point of water (above 80°C in Figure 4.1b). This is in agreement with the
criteria for formation of an enthalpic network having highly aggregated intermolecular
associations as structural knots. It has further been postulated that such aggregates are
composed of co-axial double helices on the basis of X-ray fiber diffraction and optical
rotation evidence.41
0
1
2
3
4
5
0 20 40 60 80 100 120 140 160 180
Time (min)
Lo
g (
mo
du
lus/
Pa)
-10
0
10
20
30
40
50
60
70
80
Tem
per
atu
re (
°C)
G'
G"
b
Fig. 4.1b. Storage (�) and loss (�) modulus variation as a function of temperature ( )
and time of observation for 2% agarose (scan rate: 1°C/min; frequency: 1 rad/s;
strain:0.1%; equilibration temperature: 0oC).
31
0
1
2
3
4
5
0 20 40 60 80 100
Time (min)
Lo
g (
mod
ulu
s/P
a) a
0
10
20
30
40
Tem
per
atu
re (
°C)
Fig. 4.2 Storage (�) and loss (�) modulus variation as a function of temperature ( ) and
time of observation for a) 20% gelatin and b) 2% agarose (scan rate: 1°C/min; frequency:
1 rad/s; strain: 0.1%; equilibration temperature: 25oC).
0
1
2
3
4
5
0 20 40 60 80 100
Time (min)
Lo
g (
mod
ulu
s/P
a)
aa
0
10
20
30
40
50
Tem
pera
ture
(°C
)
G’
G”
G’
G”
b
a
32
Figs. 4.2 a & b show the cooling profiles of single gelatin and agarose gel systems cooled
down to and equilibrated 25oC. Akin to the cooling traces observed at 0oC, at 25oC
gelation of samples is followed by an isothermal development of gel moduli.
Furthermore, a comparison of the results obtained by cooling down the gels to the two
designated temperatures clearly indicates that the values of storage modulus obtained
exhibit a significant dependence on the equilibration/curing temperature chosen. Thus,
for the same concentration of a gel, the lower the curing temperature, the higher the
values of moduli obtained. However, the temperature at which gel formation occurs
remains the same as there is no change in the scan rate used while cooling or in the
concentration of the gel. This difference in gel modulus values can be attributed to the
fact that further development of crosslinks is promoted at lower temperatures giving
harder gels. Additionally, the differences in gelation patterns and mechanisms of agarose
and gelatin are made explicit through these results. It is evident that the extent of
development of gel modulus as a function of time is much more for gelatin than for
agarose. Triple helix formation in gelatin gels which is promoted with progression in time
is the primary reason for the same. Also noteworthy is the observation that the
concentration of agarose needed is very little in comparison to gelatin in order to obtain
gels of similar strength.
By and large, the approach used in previous investigations of biopolymer mixed
gels constituted guidelines used in the attempt to model the tertiary systems, the crux of
this study. In doing so, the current section reports results obtained from work expanded to
include a wide range of experimental temperatures and polymer concentrations in
preparation for studies of the composite gels. The first step to laying the foundation for
33
carrying out modeling based on polymer blending laws is relating the overall mechanical
properties to those of the constituent phases. Accordingly, Figure 4.3 depicts the
concentration dependence of storage modulus for agarose and gelatin using values that
were obtained at the end of the isothermal runs for the entire range at which
measurements were made. Resultant calibration curves allow tracing the shear modulus
as a continuous function of concentration for each of the component polymers. The
double logarithmic depiction of parameters results in empirical straight-line fits that can
be used for interpolation between measured values of G' at different concentrations.
Experimental values of G’ obtained will be later on be validated with values obtained
from these plots for a range of hypothesized concentrations and phase volumes.
1.5
2.5
3.5
4.5
5.5
0.0 0.3 0.6 0.9 1.2 1.5 1.8
Log (Polymer Concentration)
Log (G
' / P
a)
agarose
gelatin
Fig. 4.3. Calibration curves of storage modulus as a function of polymer concentration
for agarose (triangles) and gelatin (circles) at 0 and 25°C (closed and open symbols,
respectively) from data recorded at the end of the isothermal runs in Figure 4.1 and 4.2.
R2= 0.9934
R2= 0.9868
R2= 0.9843
R2= 0.9887
34
The lowest experimentally-accessible concentration of the gelatin gel (1.25% at 0°C
in Figure 4.3) exhibits moduli that form a trajectory towards the minimum critical gelling
concentration of the protein. The results obtained are quite precise in indicating clearly
that at lower concentrations of gelatin, associations are intramolecular rather than
intermolecular, which is what is expected for gelatin gels. Values of ‘effective
concentration’ (utilized in subsequent modeling) obtained from these plots fall within the
higher concentration regime, where very acceptable linear fits have been obtained as
indicated by the R-square values. The gap in modulus development of gelatin gels at 0
and 25°C, as contrasted to the closely running patterns of the agarose gels using an
identical isochronic/isothermal routine, indicates that suitable conditions of temperature
and concentration have been identified for the subsequent investigation of
thermodynamic incompatibility in the tertiary system.
Further, rationalization of the changes in mechanical characteristics observed in the
tertiary system requires mapping out the cooling-isothermal-heating profile of the lipid
phase. This is reproduced in Figure 4.4 for Dalda Vanaspati lipid that was control cooled
(1°C/min) to 0°C to induce structure formation. Compared with the corresponding
profiles of the two macromolecules in Figure 4.1(a–b), the setting (30.8°C) and melting
(37.0°C) temperatures are close to the gelatin gel reflecting the melt-in-the-mouth
property of both materials. However, the general form of modulus development during
the experimental routine is quite distinct, achieving at 0°C storage modulus values in
excess of 108 Pa and tan δ ratios of the order of 0.20.
In contrast to high-solid mixtures of industrial polysaccharides and sugars, the
multifold increment in shear modulus values of which is mainly attributed to the
35
phenomenon of glass transition occurring on lowering the temperature, the present
development in viscoelasticity relates instead to the molecular structure of the
triacylglycerol molecules, which make up the network of the lipid phase.43 Further, it
appears that the microstructure of triacylglycerols, in terms of the shape, size and spatial
distribution patterns of the crystal,44
has been evolving during the isothermal sweep at
0°C from α to β' polymorphic forms hence affecting the mechanical fingerprint in Figure
4.4. By contrast, the complex dynamic viscosity (η*) of soybean oil remains unaltered
throughout the experimentally accessible shear rate, i.e., exhibiting Newtonian behavior
of ~0.07 mPa s at 25°C as seen in figure 4.5.
1
2
3
4
5
6
7
8
9
0 20 40 60 80 100 120 140 160
Time (min)
Log (
modulu
s/P
a)
-10
0
10
20
30
40
50
Tem
per
ature
(°C
)
G'
G"
Fig. 4.4. Storage (�) and loss (�) modulus variation as a function of temperature ( )
and time of observation for Dalda Vanaspati lipid (scan rate: 1°C/min; frequency: 1 rad/s;
strain: 0.01%).
36
0.01
0.1
1
0.1 1 10 100
Rate (s-1
)
|η|
|η|
|η|
|η|*
(P
a.s
)
Fig. 4.5. Complex viscosity variation as a function of shear rate for soybean oil at 25oC.
4.2 Experimental Observations on Mixed Systems of Gelatin, Agarose and a Lipid
Phase
The preceding section described the small-deformation oscillatory behaviour of
individual materials in preparation for work in binary and tertiary composite gels. Mixed
gel networks were prepared by keeping the agarose concentration constant at 1% and
increasing that of gelatin to 30% in the formulation. Further composite materials were
prepared by holding the agarose and gelatin concentration at 1% and 10%, respectively,
and increasing the lipid phase to 30% at concentration steps of 5%. The final setting
temperatures were 0 and 25°C, and representative cases of the results obtained will be
discussed presently.
Figure 4.6a shows the melting profile of 1% agarose plus 5% gelatin gel following
controlled heating at 1°C/min. Based on the mechanical fingerprints of the single
components from Figure 4.1(a-b), it appears that disintegration of the gelatin structure is
accompanied by a reduction in composite moduli, but the gel remains intact until the
37
higher temperature range associated with the melting of the agarose network. This
behaviour is consistent with the postulate that rapidly gelling agarose forms a continuous
network and melting of gelatin above 37°C weakens the gel but does not destroy it.
Incorporation of the lipid phase in the form of discontinuous inclusions does not affect
the temperature band of phase transitions since its early melting profile coincides with
that of gelatin (20% lipid-phase inclusion in the sample of Figure 4.6b). In the same
figure, there is a very evident weakening of the overall mechanical properties with
melting of the 15% gelatin gel but agarose still forms a continuous phase. Following
melting of the gelatin helices, the values of storage modulus increase and reach a plateau
between 50 and 55°C. This observation is reproducible and may be attributable to
progressive migration of non-gelling concentrations of agarose from the gelatin phase to
join the existing network of the polysaccharide that acts as a nucleus towards partial
reactivation of the process of network formation. Further heating results in gradual
disintegration of the three-dimensional agarose structure. At the highest experimental
concentration of gelatin (30% in Figure 4.6c), the composite network melts out
completely over the temperature range of the protein, an outcome which argues that
agarose is now dispersed as a discontinuous filler in the gelatin continuous matrix.
38
Fig. 4.6. Heating profiles of storage (�) and loss (�) modulus for a) 1% agarose plus 5%
gelatin, b) 1% agarose plus 15.0% gelatin plus 20.0% Dalda Vanaspati lipid and c) 1%
agarose plus 30% gelatin (scan rate: 1°C/min; frequency: 1 rad/s; strain: 0.1%).
1
2
3
4
5
0 10 20 30 40 50 60 70 80
Temperature (°C)
Lo
g (
mo
du
lus/
Pa)
G'
G"
a
0
1
2
3
4
5
6
0 10 20 30 40 50 60 70 80 90
Temperature (°C)
Lo
g (
mo
du
lus/
Pa)
G'
G"
b
0
1
2
3
4
5
6
0 10 20 30 40 50 60 70
Temperature (°C)
Lo
g (
mo
du
lus/
Pa)
G'
G"
c
39
Further evidence of phase separation in the agarose/gelatin/lipid mixture is afforded
by differential scanning calorimetry (DSC), and setting exotherms from cooling scans at
1°C/min are depicted in Figure 4.7(a-f). The DSC scans reproduced as parts (a) and (b)
for single agarose and gelatin samples exhibit an identical sequence of structure
formation as for the mechanical spectra in Figure 4.1(a-b). A relatively sharp peak
emphasizes the co-operative conformational transition of the polysaccharide segments, as
compared to the on-going structure development in the gelatin network. Thermograms of
the lipids used in this study record two distinct molecular processes: i) an extensive
thermal event in terms of heat flow (and ∆H change) for Dalda Vanaspati lipid (Figure
4.7d), as compared to that of agarose and gelatin in Figures 4.7a and 4.7b; this should
lead to the formation of crystal aggregates that interact to form a colloidal fat crystals
network44
and ii) a featureless baseline of heat flow that reflects the liquid-like nature of
soybean oil throughout the temperature range examined (Figure 4.7e).
Upon mixing, the temperature sequence and overall peak form of each transition
was maintained both for the binary agarose/gelatin mixture in Figure 4.7c and the tertiary
system of agarose/gelatin/lipid in Figure 4.7f. It is noted that the temperature band of the
agarose gel in the mixtures is shifted to higher temperatures – compare parts (a) and (f),
in accordance with the increased thermal stability of concentrated preparations of the
polysaccharide. The obvious interpretation of the observed behaviour is that each
polymeric component forms its own network within the mixed gel and the non-
interacting phases support the lipid component in the tertiary matrix. In contrast,
heterotypic associations of polymeric segments A and B usually manifest themselves by
40
distorting the peaks of the individual gels and generating a new thermal event in the DSC
spectrum (A + B).45
0.75
0.77
0.79
0.81
0.83
0.85
0 10 20 30 40 50
Temperature (oC)
Heat fl
ow
(m
W)
a
0.0
0.4
0.8
1.2
1.6
2.0
0 10 20 30 40 50
Temperature (oC)
Heat Flo
w (m
W)
d
0.74
0.76
0.78
0.80
0.82
0.84
0 10 20 30 40 50
Temperature (oC)
Heat
flo
w (m
W)
b
0.10
0.15
0.20
0.25
0.30
0 10 20 30 40 50
Temperature (oC)
Hea
t Flo
w (
mW
)
e
0.69
0.71
0.73
0.75
0.77
0.79
0.81
0 10 20 30 40 50
Temperature (oC)
Hea
t fl
ow
(m
W)
c
0.72
0.74
0.76
0.78
0.80
0.82
0.84
0 10 20 30 40 50
Temperature (oC)
Heat
flo
w (
mW
)
f
Fig. 4.7. DSC exotherms for a) 2% agarose, b) 15% gelatin, c) 2% agarose plus 15%
gelatin, d) Dalda Vanaspati lipid, e) soybean oil, and f) 2% agarose plus 15% gelatin plus
2.5% Dalda Vanaspati lipid (scan rate: 1°C/min).
41
As a final confirmation of the phase-separated morphology of our system, the
distribution of the three components within the composite network was visualized
directly using scanning electron microscopy (SEM). Figure 4.8(a-h) illustrates SEM
micrographs obtained for single and mixed preparations at 1000x magnification. The
agarose and gelatin components appear as dark and white areas, respectively, because of
their distinct phase-shifts. At low levels of gelatin addition - e.g., 2.5% in part (a),
agarose forms a continuous featureless background, an observation which is congruent
with the melting profile in Figure 4.6(a). The polysaccharide pattern is interrupted by
well formed structures of the protein, which include several “pores” with a broad size
distribution. Increasing concentration of gelatin from part (b) to (f) results in regularly
organized and densely packed gelatin strands whose “honeycomb” structures at 30%
solids form the continuous phase in the mixture in accordance with the melting profile in
Figure 4.6(c). The last two micrographs were obtained for tertiary mixtures of 1%
agarose plus 10% gelatin in the presence of 15% Dalda Vanaspati lipid or 10% soybean
oil; parts (g) and (h), respectively. Clearly, blending of materials results in isotropic
formation of structures where the polymeric matrix supports the discontinuous inclusions
of the lipid phase. These spheroidal inclusions will serve as the basis for modeling the
mechanical properties of the tertiary gel containing rigid (solid lipid) or soft (soybean oil)
filler particles.
42
a b
c d
e f
g h
Fig. 4.8. Scanning electron microscopy images of a) 1% agarose plus 2.5% gelatin, b) 1%
agarose plus 5% gelatin, c) 1% agarose plus 10% gelatin, d) 1% agarose plus 15%
gelatin, e) 1% agarose plus 25% gelatin, f) 1% agarose plus 30% gelatin, g) 1% agarose
plus 10% gelatin plus 15% Dalda Vanaspati lipid, and h) 1% agarose plus 10% gelatin
plus 10% soybean oil (scale bar is 100 µm).
43
Having reported qualitatively on the incompatibility phenomena observed in
mixtures of agarose/gelatin/lipid, the present section summarizes in the form of master
curves of shear storage modulus the concentration effect of all systems at the setting
temperatures of choice. As shown in Figure 4.9(a-b), the concentration-dependence of gel
modulus for the polymeric mixture is enveloped by the upper and lower boundaries of
agarose and gelatin, respectively. The differential of twenty five degrees centigrade in
setting temperatures does not affect the rigidity of individual polysaccharide networks,
which reach values close to 105 Pa at 3% polymer. In contrast, the corresponding
modulus-concentration relationships for gelatin are temperature driven, with moduli
varying between one and two orders of magnitude as the concentration of the protein
decreases from 30 to 5% in the single gel.
The variation of storage modulus with temperature that accompanies the setting of
single gelatin networks creates, in conjunction with the agarose data, a broad envelope of
rigidity at 25°C that affords further insights into the phase topology of the binary
mixtures, as assisted by mechanistic modeling (impending part of the discussion). It is
illustrated clearly in Figure 4.9(a-b) that the composition dependence of storage modulus
progresses from the polysaccharide to protein dominated topology with increasing gelatin
concentrations in the binary blend. This is in accordance with the melting profiles of
selected composite gels in Figures 4.6a and 4.6c. Furthermore, there is an immediate
increase in the modulus of the binary gel upon addition of gelatin, an outcome which
indicates that the two components have phase separated during gel formation.
44
2.0
3.0
4.0
5.0
6.0
0 5 10 15 20 25 30 35
Concentration (% w/w)
Log (
G' /
Pa)
a
2.0
3.0
4.0
5.0
6.0
0 5 10 15 20 25 30 35
Concentration (% w/w)
Log (
G' /
Pa)
b
Fig. 4.9. Experimental storage modulus data obtained at a) 0oC and b) 25°C for agarose
gels (�), gelatin gels (�), and increasing levels of gelatin (�) in binary mixtures that
hold the concentration of agarose constant at 1% solids; the storage modulus of selected
mixtures quenched to 25°C is also included for comparison (�).
45
This behaviour is distinct from earlier observations on the agarose/gelatin system
found in the literature where the storage modulus falls considerably as gelatin is added,5
and the gelatin/maltodextrin mixture where there is a similar reduction in rigidity with
initial addition of the carbohydrate to the protein gel whose concentration remains fixed
at 2% solids.17
The fall in gel modulus in relation to the value appropriate to the parent
agarose or gelatin sample was rationalized on the basis of a binary (X and Y) but single-
phase solution. Upon cooling, biopolymer X sets at higher temperatures than Y,
effectively forming a gel across the whole system at the nominal (original) concentration,
and then is taken to higher concentrations (process of deswelling) by partial removal of
water as the second slower component also gels. The final modulus of this composite gel
will be the same as that of the gel that has already phase separated during gelation only if
the crosslinks of biopolymer X are sufficiently labile to re-adjust to a new equilibrium
condition.46
However, if the crosslinks are effectively permanent within the experimental
constraints, the values of shear modulus are expected to be much lower than for the
phase-separated gel. It appears that quench cooling for the gelatin/agarose and
gelatin/maltodextrin samples reported in earlier studies promotes ordering-aggregation
leading to gelation that arrests the process of phase separation between the two
components. In both cases, classical network-deswelling theory is operational giving a
very much smaller difference in modulus than in gels formed directly at the final
(effective) concentrations.47
Slow cooling at a rate of 1°C/min, on the other hand,
promotes extensive phase separation and a continuous increase in gel moduli of the
present system shown in Figure 4.9(a-b). The discussion on the cooling-rate effect has
also been checked experimentally by quench cooling selected mixtures of agarose and
46
gelatin to 25°C. Upon gelation, these produce reduced values of storage modulus, as
compared to the slowly cooled counterparts (Figure 4.9b), thus arguing for the formation
of deswelled composite systems.
Having summarized the phase behaviour of the binary system, Figure 4.10
reproduces corresponding patterns recorded for the small-deformation properties of the
tertiary mixture. This was prepared by holding the levels of agarose and gelatin at 1% and
10%, respectively, and varying those of the lipid phase from 5 to 30% in the formulation.
Experimental moduli generated following addition of Dalda Vanaspati lipid and setting at
0°C exhibit an upward incline consistent with a rigid-filler effect on the polymeric
matrix. The modulus-concentration relationship taken in relation to soybean oil at the
setting temperature of 25ºC exhibits a reverse “mirror-image” behaviour evident from the
negative slope in fig. 4.10 which is what is anticipated from the inclusion of soft-filler
particles in the gel. The macroscopic properties of the solid/semi-solid emulsions and
those of the binary aqueous gels will be considered at length in the quantitative treatise of
the next section.
47
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0 5 10 15 20 25 30 35
Concentration of the Lipid Phase (% w/w)
Log (
G' /
Pa)
Fig. 4.10. Experimental storage modulus data obtained at 0°C for 1% agarose plus 10%
gelatin with increasing concentrations of the Dalda Vanaspati lipid phase (�), and at
25°C for 1% agarose plus 10% gelatin with increasing concentrations of the soybean oil
phase (�); solid line fits reflect modeling using theoretical equations of rigid and soft-
filler composite materials.
48
4.3 Quantitative Analysis of Mechanical Functions in Support of the Phase
Topology of the Agarose/Gelatin/Lipid Mixture
Pioneering research carried out in 1983 studied properties of thermally induced gel
mixtures containing agarose and gelatin in an attempt to construct a theoretical model
using ideas of solvent partition between components as also of upper and lower bound
limits for G’ v/s composition behaviour.4 In contrast to polymer composites, for which
the blending laws were first developed, for most biopolymer composites, the presence of
water as a third component introduces an added complication. This issue was addressed
by introducing a relative affinity parameter ‘p’ which helps determine the ‘solvent
avidity’ or ‘water holding capacity’ of the biopolymers present. Unlike for pure polymer
composites, the phase volumes in aqueous binary composites depend on both, the amount
of polymer present and the manner in which the solvent partitions itself. The ‘p-factor’
thus defines the phase volumes and hence the effective concentration of each polymer
within its own phase.
Concepts of nominal and effective concentrations initially suggested an equilibrium
treatment approach. It was later debated that the final phase volumes may be under
kinetic rather than equilibrium control due to potential ‘freezing’ of the system by
network formation of one or both polymers hence eliminating re-partition of the solvent.
Subsequently, density differences between the phases were accounted for, a parameter
overlooked in previous attempts. Direct contribution to phase volume in terms of
polymeric weight was no longer neglected deeming the proposal of equating phase
volume to the volume of water in each phase by previous researchers redundant. Thus,
there has been a clear progression in modeling biopolymer systems since first attempted.
49
In doing so, two approaches, with the laws of Takayanagi as their premise, have been
commonly used and often referred to in the literature: i) concentration-dependence of gel
moduli, and ii) Network de-swelling.
For the current piece of work, conclusions reached on the basis of qualitative
thermomechanical analysis and microscopy images are complemented in this section
using theoretical frameworks developed for composite systems in an effort to formulate
reasonable predictions as a function of material composition and experimental settings.
Starting with the agarose/gelatin sample, we established in the preceding sections that the
observed phase behaviour is due to each component forming its own network in the
binary gel. However, the possibility of bicontinuous arrangements with the individual
networks phase-separating and interpenetrating one another has not been explicitly
addressed.
It is obvious that at high concentrations of the protein, shown for example in Figure
4.6c, the polysaccharide network cannot span the entire system, since heating of the
composite gel is sufficient to eliminate long-range cohesion at the low temperatures
(about 40°C) associated with the melting of the gelatin structure. Here, gelatin forms the
continuous matrix supporting discontinuous inclusions of agarose. At low concentrations
of the protein, shown for example in Figure 4.6a, it is clear that the polysaccharide forms
a continuous network that melts at temperatures above 70°C. The evidence as to the
topology of the gelatin network, i.e., continuous or discontinuous, is less direct and
requires quantitative analysis of mixed-gel moduli in relation to those obtained for the
constituent polymers in isolation.
50
The Takayanagi blending law approach (as discussed in detail on page 12) that has
been commonly used to predict the mechanical properties of biopolymer co–gels gives
isostrain and isostress conditions, obtained using the following additive formulae:48
G'c = φx G'x + φy G'y ……(4.1)
G'c = (φx / G'x + φy / G'y)-1
……(4.2)
where, G'c is the storage modulus of the composite, G'x and G'y are the moduli of the
constituent phases (X and Y), with φx and φy being their respective phase volumes (φx +
φy = 1). According to this approach, when the supporting phase is stronger than the
discontinuous inclusions equation (4.1) is applicable (isostrain conditions and upper
bound behaviour), whereas in the converse situation with the dispersed filler being
stronger than the continuous matrix equation (4.2) should be used (isostress case and
lower bound predictions).
Application of blending laws to the agarose/gelatin system, an example of which is
reproduced in Figure 4.11a, yields inconclusive results. The family of calculated bounds
is for mixtures gelled at 25°C but similar patterns are recorded for the zero-degree
counterparts. This type of plotting predicts the distribution of solvent between the two
biopolymer phases by calculating, via the calibration curves in Figure 4.3, the values of
storage modulus on shear for all possible distributions of solvent in the agarose phase
(Sagar) and finding which ones matched the experimental values (G'exp) of the mixture. A
detailed treatise on the ins-and-outs of this semi-theoretical framework has been
published earlier and readers are referred to this, which includes a discussion of the
requirement for complete phase separation in the gel.18,1
51
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Sagar
Lo
g G
' (P
a)
G' agar
G' gel
G' exp
G' U(agar)
G' L(gel)
G' U(gel)
G' L(agar)
a
Fig. 4.11a. Modeling the phase topology of the agarose/gelatin gel at 25°C using the
isostrain and isostress blending laws for a binary sample comprising 1% agarose plus
10% gelatin. Upper and lower bounds are denoted by solid lines as G'U and G'L
respectively. (gel) and (agar) stand for the gelatin and agarose phases respectively. G'exp
(---)is the experimental composite gel modulus. Dashed lines denote the derived moduli
for individual agarose and gelatin gels. Sagar is the solvent fraction in the agarose phase.
When the experimental modulus for the composite gel of 1% agarose plus 10%
gelatin is plotted in Figure 4.11a, it intersects both lower bounds of the gelatin continuous
52
(G'L(gel)) and agarose continuous (G'L(agar)) topologies. From the melting-temperature
evidence presented in Figure 4.6a (and Figure 4.6b) for mixtures of low and intermediate
gelatin concentrations, the former topology is not possible. Similarly, the arrangement of
an agarose supporting matrix claiming a disproportionately large volume of solvent in its
phase (Sagar = 0.95), thus rendering the discontinuous inclusions of gelatin effectively
dehydrated, is physically unrealistic. In conclusion, the analysis of isostrain and isostress
blending laws fails on both counts to rationalize the relationship between mixed gel
moduli and their components.
Direct visualisation of the distribution of the two components within the mixed gel
network, using scanning electron microscopy in Figure 4.8, suggests that gelatin forms a
continuous phase alongside that of agarose and this molecular architecture will be
pursued next. In doing so, we shall utilise a theoretical formula first introduced by Davies
to relate the overall modulus of the isotropic composite to the moduli and phase volumes
of the individual networks arranged in a bicontinuous manner:
G'c1/5
= φx G'x1/5
+ φy G'y1/5
……(4.3)
Details for the derivation of equation (4.3) can be found in the literature, which points out
that the first application was on the mechanical and dielectric properties of condensed
synthetic composites with a macroscopically homogeneous consistency.23,49 In hydrated
biogels, it predicted the rigidity modulus of a blend containing two continuous phases of
iota and kappa carrageenan.26
53
3.40
3.45
3.50
3.55
3.60
3.65
3.70
3.75
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Sagar
Lo
g G
' (P
a)
b
Fig. 4.11b. Modeling the phase topology of the agarose/gelatin gel at 25°C using the
Davies blending law for bicontinuous composite gels of 1% agarose plus 1.25, 2.5, 3.75,
5.0, 6.25, 7.5, and 10% gelatin (dashed lines of corresponding experimental data from
bottom to top of the graph, with solid lines from right to left reflecting calculated
predictions for the same order of binary gels).
Figure 4.11b reproduces results of this analysis for the agarose gel in the presence
of low concentrations of gelatin studied (1.25 to 10%). Gratifyingly, the observed values
of co-gel moduli intersect the appropriate predictions for each composition, and
intersected predictions shift regularly to lower volumes of solvent in the agarose phase
with increasing gelatin content in the mixture. To allow for meaningful comparisons of
54
the solvent partition in our mixtures that incorporate relatively different levels of
polymer, the “solvent-avidity” factor, p, was utilized. This was introduced in the
following format as an empirical means of accounting for the composition of the system:4
p = (Sagar/cagar)/(Sgel/cgel) ......(4.4)
where, the parameter “S” stands for the amount of solvent kept in each phase, and cagar
and cgel are the nominal (original) concentrations of the two macromolecules in
preparations. Clearly, p values contrast the level of solvent found within phases per unit
of the initial concentration of the corresponding polymer.
Intersection data in Figure 4.11b were replotted as a function of the solvent-avidity
concept to yield Figure 4.11c. Estimates of the p factor indicate a smooth (linear)
progression on the relative affinities of the two polymeric networks for solvent, which
works in gelatin’s favour with increasing concentrations of the protein in samples. This
variation in p values is distinct from previous estimates in composite gels that exhibited
isostrain or isostress phase behavior according to the blending-law equations (4.1) and
(4.2), and classic phase inversion from one continuous matrix to another.25,48
Thus, it was
found that true phase inversion in biomaterials, due to increasing concentration of the
second (Y) component, resulted in two distinct clusters of p values, one for the X–
continuous composite and the other for the Y–continuous composite following phase
inversion. The lack of any systematic variation of p with polymer concentration before
and after phase inversion, and the magnitude of p values in each cluster of points
reflected fundamental differences in the interplay between kinetics of ordering/gelation
and microscopic phase separation in those materials.
55
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 1 2 3 4 5 6 7 8 9 10 11
Gelatin (%)
p f
acto
r
c
Fig. 4.11c. Modeling the phase topology of the agarose/gelatin gel at 25°C using
predictions of the “p factor” based on the bicontinuous blending law for the composite
gels of Figure 4.11b.
Having argued for bicontinuity in the geometrical arrangement of the constituent
networks in binary gels with a low-gelatin content, the final stage of the analysis sought a
valid interpretation of the phase topology in tertiary mixtures. As mentioned in Figures
4.8(g) and 4.8(h), the drive to reduce interfacial tension between the aqueous and lipid
phases resulted in gelled matrices dispersed (isotropically) with spheroidal inclusions.
Furthermore, mechanical data in Figures 4.4 and 4.5 argue that the solid-like lipid phase
and the soybean oil inclusions should be in the form of rigid and “soft” filler particles,
respectively. The modulus of the dispersed phase divided by the modulus of the aqueous
56
phase yields values of the order of thousands for Dalda Vanaspati lipid. This modulus
ratio is roughly the correct indicator for synthetic polyblends and block polymers whose
viscoelastic characteristics can be rationalized with the theoretical advance achieved by
Lewis and Nielsen. The analytical expression of this framework of thought is given as
follows:50-52
G' / G'1 = (1 + ABφ2) / (1 - Bψφ2) ……(4.5)
B = [(G'2 / G'1) - 1] / [(G'2 / G'1) + A] ……(4.6)
ψ = 1 + [(1 - φm) / φm2
]φ2 ……(4.7)
G'1 / G' = (1 + AiBiφ2) / (1 - Biψφ2) ……(4.8)
Bi = [(G'1 / G'2) - 1] / [(G'1 / G'2) + Ai] ……(4.9)
where, in equations (4.5) to (4.7), G', G'1 and G'2 are the shear moduli of the composite,
continuous phase of the “elastomer” and the rigid filler, and φ2 is the phase volume of the
said filler. The framework takes into account the shape of the filler via the Einstein
coefficient (A), which is very sensitive to the morphology of the composite. A further
level of refinement was achieved by considering the concept of maximum packing
fraction (φm) of the filler phase.
Figure 4.10 illustrates the experimental moduli of the tertiary preparation
incorporating 1% agarose, 10% gelatin and increasing levels of lipid from 5 to 30%, and
the standard of agreement obtained using equations (4.5) to (4.7). Experimental moduli of
the aqueous and lipid phase for samples subjected to a sixty minutes isothermal run at
0°C are known, as is known the phase volume of Dalda Vanaspati lipid. For the purpose
of modeling, the Einstein coefficient was held constant at the theoretical value for
dispersed spheres (A = 1.5), hence the only adjustable parameter of the fit is the
57
maximum packing fraction of the dispersed phase. The latter is a rather difficult
parameter to estimate experimentally but, for each phase volume of the suspended
particles, very acceptable fits have been obtained for an φm value of 0.48. This
corresponds to loose packing of spheroids,53
which should reflect the low density
achievable by a random collection of extremely hard lipid particles within a soft
polymeric matrix.
In the opposite system of soybean oil inclusions, the continuous matrix is the more
rigid one and its “dilution” with liquid-like additions results in weakening of the overall
rigidity. Modeling may be attempted in such systems by inverting equations (4.5) and
(4.6) to produce equations (4.8) and (4.9), with the subscript 2 still referring to the
dispersed phase.50
In this case, the Einstein constant is now given as Ai, where Ai = 1/A.
Modeling for each phase volume of the dispersed spheres is implemented presently using
a maximum packing fraction of 0.64. Successful use of the notion of random close
packing in the descending modulus patterns of Figure 4.10 reflects the high density
achieved by the partial agglomeration (fusion) of the malleable oil droplets in the
polymeric matrix.
58
CHAPTER 5: CONCLUSIONS
In several earlier investigations of binary biopolymer mixtures where the
concentration of the rapidly gelling component was held fixed, gel moduli of the
composite showed a marked decrease with a systematic increase in the content of the
second component, an observation that was rationalized on the basis of the network de-
swelling theory. In this piece of work, we have stressed the antagonistic effect between
ordering/gelation and phase separation in the agarose/gelatin system. Thus low cooling
rates enhance phase separation leading to reinforcement in composite moduli in relation
to that of the faster-gelling polymer (agarose) at its nominal concentration. It was further
documented that the overall modulus of the phase separated gel could be related to the
moduli of the individual networks and their phase volumes by a bicontinuity blending
law. This outcome facilitated prediction of an unexpected distribution of solvent between
the two continuous phases of agarose and gelatin. Understanding the effective elastic
functions of an isotropic two-phase mixture led to the desire to control the functionality
of a complex biomaterial incorporating a lipid phase. Modeling of the tertiary system was
achieved by using ideas that relate microscopic morphology to the mechanical functions
of synthetic-polymer polyblends and may pave the way for tailor-made industrial
applications.
59
CHAPTER 6: SUGGESTIONS FOR FUTURE WORK
Work that has not been addressed explicitly in this piece of work is proposed as
potential research areas of interest for the future. The first suggestion thereof would be to
conduct phase diagram studies for the gelatin-agarose system. A phase diagram helps
ascertain the state of phase separation of mixed biopolymer solutions, an observation that
varies with levels of compatibility between the polymer components as well as the
polymer solvent affinity.54
Such studies could help assess the presence of cloudiness in
the solution and consequently define ‘cloud point curve’ or ‘binodal’ between the
monophasic and biphasic systems. This in turn could assist in better understanding the
nature of the gel phases.
The next suggestion is to carry out Isothermal/annealing studies at different
temperatures for agarose gels as well as mixed systems in order to identify the nature of
phenomena taking place on melting the gels. This will help elucidate the origin of the
anomalous melting profile. The proposition that migration and subsequent nucleation of
agarose pockets leads to a slight reinforcement of gel moduli almost immediately
following the melting of gelatin needs to be verified. For this, tempering gels at various
temperatures could help narrow down and eventually pin point causative factors for the
unusual phenomenon observed.
Finally, it is suggested that Confocal Laser Scanning Microscopy of mixed gel
systems be attempted wherein visualization of 3-D images could provide a direct
determination of phase volumes. Phase volume estimates obtained by modeling carried
out in this piece of work could then be verified.
60
REFERENCES
1. Kasapis S. and Al-Marhoobi I.M. (2005). Bridging the divide between the high-
and low-solid analyses in the Gelatin/k-Carrageenan mixture.
Biomacromolecules, 6, 14-23.
2. Harding S.E., Hill S.E. and Mitchell R. (1995). In: Biopolymer Mixtures edited
by Harding S.E., Hill S.E. and Mitchell R., Nottingham University Press, pp v-
viii.
3. Kasapis, S. (2008). Phase separation in biopolymer gels: A low to high-solid
exploration of structural morphology and functionality. Critical Reviews in Food
Science and Nutrition, 48, 341-359.
4. Chronakis I.S. and Kasapis S. (1995). A rheological study on the application of
carbohydrate-protein incompatibility to the development of low fat commercial
spreads. Carbohydrate Polymers, 28, pp 367-373.
5. Clark A.H., Richardson R.K., Ross-Murphy S.B. and Stubbs J.M. (1983).
Structural and mechanical properties of Agar/Gelatin co-gels. Small deformation
studies. Macromolecules, 16, 1367-1374.
6. Lundin L., Norton I.T., Foster T.J., Williams M.A.K., Hermansson A.-M. and
Bergstrom E. (2000). Phase Separation in Mixed Biopolymer Systems. In: Gums
and Stabilizers for the Food Industry 10 edited by Williams P.A. and Phillips
G.O., The Royal Society of Chemistry, 167-180.
7. Ross Murphy S.B. (1991). Structure and Rheology of Gelatin gels: recent
progress. Polymer, 33, 12, 2622-2627.
61
8. Michon C., Cuvelier G., Relkin P. and Launay B. (1997). Influence of thermal
history on the stability of gelatin gels. International Journal of Biological
Macromolecules, 20, 259-264.
9. Tako M. and Nakamura S. (1988). Gelation mechanism of Agarose.
Carbohydrate Research, 180, 277-284.
10. Ross-Murphy S.B. (1995). Small deformation rheological behaviour of
biopolymer mixtures. In: Biopolymer Mixtures edited by Harding S.E., Hill S.E.
and Mitchell R., Nottingham University Press, 85-98.
11. Richardson R.K. and Kasapis S. (1998). Rheological methods in the
characteristics of food biopolymers. In: Instrumental methods in food and
beverage analysis edited by Wetzel D. and Charalambous G., Elsevier Science
B.V., 1- 48.
12. Morris, E.R. (1995). Polysaccharide synergism: more questions than answers?
In "Biopolymer Mixtures" (S.E. Harding, S.E. Hill and J.R. Mitchell, Eds.),
pp. 247-288. Nottingham University Press, Nottingham, UK.
13. Dea, I.C.M., Morris, E.R., Rees, D.A., Welsh, E.J., Barnes, H.A., and Price, J.
(1977). Associations of like and unlike polysaccharides: mechanism and
specificity in galactomannans, interacting bacterial polysaccharides, and related
systems. Carbohydrate Research, 57, 249-272.
14. Morris, E.R. (1998). Segregative interactions in biopolymer co-gels. In
"Phase/State Transitions in Foods: Chemical, Structural, and Rheological
Changes" (M.A. Rao and R.W. Hartel, Eds.), pp. 159-186. Marcel Dekker, New
York.
62
15. Kasapis S., Morris E.R., Norton I.T. and Brown R.T. (1993). Phase equilibria and
gelation in gelatin/maltodextrin systems- Part I: gelation of individual
components. Carbohydrate Polymers, 21, 243-248.
16. Kasapis S., Morris E.R., Norton I.T. and Brown R.T. (1993). Phase equilibria and
gelation in gelatin/maltodextrin systems- Part II: polymer incompatibility in
solution. Carbohydrate Polymers, 21, 249-259.
17. Kasapis S., Morris E.R., Norton I.T. and Brown R.T. (1993). Phase equilibria and
gelation in gelatin/maltodextrin systems- Part III: phase separation in mixed gels.
Carbohydrate Polymers, 21, 261-268.
18. Kasapis S., Morris E.R., Norton I.T. and Brown R.T. (1993). Phase equilibria and
gelation in gelatin/maltodextrin systems- Part IV: composition dependence of
mixed-gel moduli. Carbohydrate Polymers, 21, 269-276.
19. Morris E.R. (1992). The effect of solvent partition on the mechanical properties of
biphasic biopolymer gels: an approximate theoretical treatment. Carbohydrate
Polymers, 17, 65-70.
20. Kasapis S. (1995). Phase Separation in Hydrocolloid Gels. In: Biopolymer
Mixtures edited by Harding S.E., Hill S.E. and Mitchell R., Nottingham
University Press, 193-224.
21. Giraudier, S., Hellio, D., Djabourov, M. and Larreta-Garde, V. (2004). Influence
of weak and covalent bonds on formation and hydrolysis of gelatin networks.
Biomacromolecules, 5, 1662-1666.
22. Alevisopoulos, S. and Kasapis, S. (1999). Molecular weight effects on the
gelatin/maltodextrin gel. Carbohydrate Polymers, 40, 83-87.
63
23. Norton, I.T. and Frith, W.J. (2001). Microstructure design in mixed biopolymer
composites. Food Hydrocolloids, 15, 543-553.
24. Davies, W.E.A. (1971). The theory of elastic composite materials. J. Phys. D:
Appl. Phys., 4, 1325-1339.
25. Morris E.R. (2000). Rheology of Biopolymer co-gels. In: Hydrocolloids-Part 2
edited by Nishinari K., Elsevier Science, 135-146.
26. Gilsenan P.M., Richardson R.K. and Morris E.R. (2003). Associative and
segregative interactions between gelatin and low-methoxy pectin. Part 3.
Quantitative analysis of co-gel moduli. Food Hydrocolloids, 17, 751-761.
27. Piculell, L., Nilsson, S. and Muhrbeck. P. (1992). Effect of small amounts of
kappa-carrageenan on the rheology of aqueous iota-carrageenan. Carbohydrate
Polymers, 18, 199–208.
28. McEvoy H., Ross-Murphy S.B. and Clark A.H. (1985). Large deformation and
ultimate properties of biopolymer gels: 1. Single biopolymer component systems.
Polymer, 26, 1483-1492.
29. Mohammed Z.H., Hember M.W.N., Richardson R.K. and Morris E.R. (1998).
Application of polymer blending laws to composite gels of agarose and
crosslinked waxy maize starch. Carbohydrate Polymers, 36, 27–36.
30. Mohammed Z.H., Hember M.W.N., Richardson R.K. and Morris E.R. (1998). Co-
gelation of agarose and waxy maize starch. Carbohydrate Polymers, 36, 37–48.
31. Burin, L., Jouppila, K., Roos, Y., Kansikas, J. and del Pilar Buera, M. (2000).
Color formation in dehydrated modified whey powder systems as affected by
compression and Tg. J. Agric. Food Chem., 48, 5263-5268.
64
32. Thomsen, M.K., Lauridsen, L., Skibsted, L.H. and Risbo, J. (2005). Temperature
effect on lactose crystallization, Maillard reactions, and lipid oxidation in whole
milk powder. J. Agric. Food Chem., 53, 7082-7090.
33. Lievonen, S.M., Laaksonen, T.J. and Roos, Y.H. (2002). Nonenzymatic browning
in food models in the vicinity of the glass transition: Effects of fructose, glucose,
and xylose as reducing sugars. J. Agric. Food Chem., 50, 7034-7041.
34. Kasapis, S. (2006). Building on the WLF/Free Volume Framework: Utilization of
the Coupling Model in the Relaxation Dynamics of the Gelatin/Cosolute System.
Biomacromolecules, 7, 1671-1678.
35. Karim, A.A. and Bhat, R. (2009). Fish gelatin: properties, challenges, and
prospects as an alternative to mammalian gelatins. Food Hydrocolloids, 23, 563-
576.
36. Amici, E., Clark, A.H., Normand, V. and Johnson, N.B. (2002). Interpenetrating
Network Formation in Agarose−κ-Carrageenan Gel Composites.
Biomacromolecules, 3, 466-474.
37. Lim, K.P. and Tan, L.P. (2009). Interaction force measurements for the design of
tissue adhesives. Acta Biomaterialia, 5, 84-92.
38. te Nijenhuis K. (1997). Introduction. Adv Polym Sci, 130, 1-12.
39. Pinhas, M.F., Blanshard, J.M.V., Derbyshire, W. and Mitchell, J.R. (1996). The
effect of water on the physicochemical and mechanical properties of gelatin.
Journal of Thermal Analysis and Calorimetry, 47, 1499-1511.
65
40. Normand, V., Lootens, D.L., Amici, E., Plucknett, K.P. and Aymard, P. (2000).
New insight into agarose gel mechanical properties. Biomacromolecules, 1, 730-
738.
41. Arnott, S., Fulmer, A., Scott, W.E., Dea, I.C.M., Moorhouse, R. & Rees, D.A.
(1974). The agarose double helix and its function in agarose gel structure. J. Mol.
Biol., 90, 269-284.
42. Clark, A.H. and Ross-Murphy, S.B. (1987). Structural and mechanical properties
of biopolymer gels. Adv Polym Sci, 83, 57-192.
43. Kasapis, S. (2008). Recent advances and future challenges in the explanation and
exploitation of the network glass transition of high sugar / biopolymer mixtures.
Critical Reviews in Food Science and Nutrition, 48, 185-203.
44. Tang, D.M. and Marangoni, A.G. (2007). Modeling the rheological properties and
structure of colloidal fat crystal networks. Trends in Food Science & Technology,
18, 474-483.
45. Fitzsimons, S.M., Tobin, J.T. and Morris, E.R. (2008). Synergistic binding of
konjac glucomannan to xanthan on mixing at room temperature. Food
Hydrocolloids, 22, 36-46.
46. Boral, S., Gupta, A.N. and Bohidar, H.B. (2006). Swelling and de-swelling
kinetics of gelatin hydrogels in ethanol-water marginal solvent. International
Journal of Biological Macromolecules, 39, 240-249.
47. Hamidi, M., Azadi, A. and Rafiei, P. (2008). Hydrogel nanoparticles in drug
delivery. Advanced Drug Delivery Reviews, 60, 1638-1649.
66
48. Kontogiorgos, V., Ritzoulis, C., Biliaderis, C.G. and Kasapis, S. (2006). Effect of
barley β-glucan concentration on the microstructural and mechanical behaviour of
sodium caseinate gels. Food Hydrocolloids, 20, 749-756.
49. Davies, W.E.A. (1971). The elastic constants of a two-phase composite material.
J. Phys. D: Appl. Phys., 4, 1176-1181.
50. Nielsen, L.E. (1974). Morphology and the elastic modulus of block polymers and
polyblends. Rheol. Acta, 13, 86-92.
51. Petersson, L. and Oksman, K. (2006). Biopolymer based nanocomposites:
Comparing layered silicates and microcrystalline cellulose as nanoreinforcement.
Composites Science and Technology, 66, 2187-2196.
52. Fornes, T.D. and Paul, D.R. (2003). Modeling properties of nylon 6 / clay
nanocomposites using composite theories. Polymer, 44, 4493-5013.
53. Torquato, S., Truskett, T.M. and Debenedetti, P.G. (2000). Is random close
packing of spheres well defined? Physical Review Letters, 84, 2064-2067.
54. Cesaro A, Cuppo F., Fabri D., and Sussich F. (1999). Thermodynamic behaviour
of mixed biopolymers in solution and in gel phase. Thermochimica Acta, 328,
143-153.
55. Brown C.R.T., Foster T.J., Norton I.T. and Underdown J. (1998). Influence of
Shear on the Microstructure of Mixed Biopolymer Systems. In: Biopolymer
Mixtures edited by Harding S.E., Hill S.E. and Mitchell R., Nottingham
University Press, 65-83.
PART II
EFFECT OF CO-SOLUTE (GLUCOSE SYRUP) ON THE STRUCTURAL
BEHAVIOUR OF AMYLOSE GELS
68
CHAPTER 1: INTRODUCTION
Starch, a common and essential component of the human diet, occurs as a storage
carbohydrate in various plant organs. It forms the primary source of carbohydrates in
human nutrition and is therefore an important ingredient of most foods. Numerous
sources of starch have been cited in literature, with that obtained from potatoes, corn and
wheat being most commonly referred to. Depending on their origin, starches differ in
their properties such as granule size, shape, distribution, composition and crystallinity.
Typically, starch occurs in nature as water insoluble, roughly spherical granules of
varying size, size distribution and shape with an ordered and partially crystalline granular
internal structure.1 Studies have revealed 70% of the mass of a starch granule as
amorphous and 30% as crystalline. Thus, starch exhibits a semi-crystalline character and
the degree of crystallinity depends on the water content. The main components of starch
are two structurally distinct polysaccharides, amylose and amylopectin. In addition to
these, small amounts of lipids and proteins are also known to be present.
On heating an aqueous suspension of starch above a characteristic temperature,
irreversible swelling of the granules occurs which is accompanied by a loss of order,
birefringence and crystallinity.1 The entire process involves sequential steps of water
diffusion into the starch granules causing them to rupture/burst followed by leaching out
of amylose into solution. On cooling this solution, an opaque elastic composite gel is
obtained, the continuous matrix of which is attributed to the gelation of amylose.1-5
Swollen granules with an amylopectin skeleton remain entrapped within the solubilized
amylose.
69
Sugars are common co-solutes in baked goods as they are known to impart an anti-
staling effect. They are employed to manipulate the glass transition temperature (Tg)
where alteration of the crystallization rate is required. Amylose crystallization on the
other hand, plays an important role in the digestibility of starch based foods as it is
known to be one of the inhibiting factors in the enzymic digestion of starch.1 The
resistance of starch to digestion can thus be determined and altered by manipulating
amylose crystallization. Further, depending on the Tg of the product in question, a wide
variety of starch based foods with different textures can be obtained. Starch based foods
are thus complex materials comprising proteins, water, sugars, lipids, amylose and
amylopectin. For the aforementioned reasons, it is of great interest to study in particular,
the interactions between amylose and sugars apart from the many others possible between
other components.
1.1 Amylose
Amylose is a linear (1-4) linked α-D-glucan.2 It behaves as a flexible coil when in
dilute, neutral, aqueous solution. Depending on the polymer concentration and molecular
weight, amylose precipitates from solution. Concentrated solutions of amylose form
opaque elastic gels on cooling down to room temperature.3 X-ray diffraction studies
exhibit resemblance to the β-type crystalline form observed for starch. Studies suggest
that amylose packs as antiparallel double-helices in a hexagonal array.
Amylose gelation has been described as amylose precipitation from solution. It has been
regarded as a partial crystallization process. Previous researchers have also termed it as
reterogradation (return to the granular state).2 Two processes contribute to the early
70
stages of gelation. Phase separation into polymer rich and polymer deficient regions
followed by development of crystalline regions within the polymer-rich regions has been
suggested as the mechanism of gelation.3 Amylose aggregation has been described as the
adoption of double-helix structures, followed by helix-helix aggregation, leading to
turbidity effects, crossed linked networks or gelation depending on the length of the
amylose chains.4
There is a discrepancy in literature with regards to the minimum concentration
required for gelation (C0) and the concentration required for coil overlap (C*)2,4,5
. Some
reports suggest that gelation can occur well below C* while others claim that gel
formation begins at concentrations above C*. The sol-gel transition takes place
sufficiently slowly for accurate measurement in the vicinity of C0. At low concentrations,
turbidity is displayed but without gelation. At high concentrations, the process of gelation
is too rapid to be described accurately.
The structural behaviour of various polymer networks in the presence of sugars has
been cited in literature.7 In doing so, the structural characteristics of different biopolymer
gels were examined at different levels (low, intermediate and high) of co-solute (sugar).
At intermediate levels of co-solute a collapse was observed in the case of all the
polysaccharide gels examined, while a reinforcement of network strength was observed
for Gelatin, a protein. Some of the gelling polysaccharides examined were agarose, κ-
carrageenan and deacylated gellan. The structural properties of amylose in the presence
of sugar have not been investigated so far. It is therefore of interest to see what pattern
amylose exhibits as a function of sugar present.
71
1.2 Glass transition in high solid systems
Literature defines glass as an amorphous solid, or an undercooled high viscosity
liquid that has lost its ability to flow.8,9 High solid materials exhibit aspects of
temperature-induced vitrification phenomena.10
The phenomenon of glass transition has
been extensively studied in amorphous synthetic polymers and the concept has been
extended to food systems. Food systems however are governed by kinetics rather than
thermodynamics as most products as well as processing procedures are non-equlilibrium
in nature.8 Time-dependent structural and mechanical properties related to the storage
stability, quality and texture in foods are better understood by applying the concept of
glass transition to food materials. When a liquid is cooled so quickly such that
crytallisation is not allowed to occur, glass is formed. The process of glass formation is
also commonly known as vitrification. Typically, amorphous or partially crystalline
materials undergo vitrification.
Glass transition involves a change in state but no change in phase of the material. It
is a second order thermodynamic transition.8,11
The glass transition temperature (Tg) is
generally regarded as the mean temperature at which the material changes from a glassy
solid to a rubbery liquid. Since the transition is a dynamic process, Tg is affected by
frequency and cooling rates.
The methods widely employed to study glass transitions in systems are calorimetry
and rheology. Depending on the method employed to determine the Tg, there are two
types of Tg, the network Tg (rheological Tg) and the calorimetric Tg. Both have the same
value for a high sugar system. For systems containing biopolymers however, the
rheological Tg is usually higher than the calorimetric Tg. Calorimetry reports changes
72
observed in heat flow (heat capacity) as a function of temperature to produce a distinct
experimental trace, the asymptotes to which demarcate the limits and the middle of the
calorimetric glass transition region. The mechanical glass transition temperature on the
other hand has been defined as “combinations of temperature and deformation frequency
for which as temperature decreases, the large scale rotational and translational motions of
enough large numbers of mobile units such as small molecules or backbone chain
segments of a macromolecule, become cooperatively immobilized during a timescale
comparable to the experimental period.” At the mechanical Tg, the free volume has been
predicted to be less than 3% of the total volume.8,12
Determination of the aforementioned ‘rheological Tg’ is usually carried out by using
the free volume theory that has been widely adopted to study glass transitions in
polymers. Free volume (vf) is the difference between the total volume (vt) and occupied
volume (vo) per mole of molecules in a material. It is that volume that is available for
large scale vibrational motions of molecules. The occupied volume includes the Van der
Waals radii as well as the volume related to molecular vibrations of the molecules.12
A popular quantitative approach to the theory of free volume is the Williams-
Landel-Ferry (WLF) model. The WLF equation, which relates viscosity to temperature, is
given by,
o2
o1T
o TTC
)T(TCa log
)η(T
η(T) log
−+
−−==
--- (2)
73
where η(T) and η(To) are viscosities at temperature T and To respectively, aT is the shift
factor, and C1 and C2 are the WLF constants which are equal to 1/2.303 fo and fo/ αf
respectively.13
The WLF equation could be derived by assuming a linear relationship between f
and T-To,
f = fo + αf (T – To) --- (3)
where f and fo are the fractional free volumes (vf/vT) at temperature T and To respectively,
and αf is thermal expansion coefficient of free volume. Substituting aT = ηoToρo/ ηTρ
into equation (1),
)Tρ
ρTlog( )
f
1
f
1(
2.303
B log oo
o
T +−=a --- (4)
where ρ and ρo are the densities at T and To. The last term is ignored by assuming a slow
temperature variation of Tρ. Equation (3) is substituted into (4) to obtain equation (2).
Rheological data such as viscosity, G’ and G” could be used to generate the shift
factors in the WLF equation. Using the method of reduced variables, also known as the
time-temperature superposition (TTS), mechanical spectra obtained at different
temperatures are horizontally superposed to generate master curves.14
TTS has also been
applied to high sugar/biopolymer mixtures.15
However, care has to be taken when
74
applying TTS as the concept assumes that all relaxation processes have the same
temperature dependence which may not be true for different relaxation mechanisms.14
The relationship between viscoelasticity of materials and chemical structure has
become of academic and technological interest in recent times. For the mechanical and
relaxation properties of amorphous synthetic polymers, empirical functions have been
shown to follow the effects of temperature, timescale of observation, molecular weight,
concentration for diluted systems, and high hydrostatic pressure.16-18
Accordingly, the
concept of free volume was argued as adequate for the rationalization of observations in
the so called “rubber-to-glass transition”.19,20
Lately, the approach has been refined by
“the coupling model”, which was introduced on the basis that intermolecular interactions
in densely packed systems are the ultimate determining factor of molecular dynamics.21,22
In future, this could potentially help overcome possible simplifications associated with
the application of the free volume theory to the entirety of the glass transition region.
In high-solid biosystems, such considerations have been dealt with rather recently
following appreciation that the glassy consistency is common in a number of natural and
man-made materials. In view of the above, interpretation of the thermal, mechanical and
spectroscopic characteristics of biological glasses and melts makes use of ideas from
polymer physics- the backbone of materials science. In several sources, the
transformation from the rubber to the glassy state is considered to be, albeit with partial
justification, in the nature of a second-order thermodynamic transition in which the
system undergoes a change in state but not in phase.23
This was then discussed in relation
to the problem of keeping the quality or shelf life of processed products.24
Vitrification
was considered as the reference point and generally, discussion of molecular properties
75
has been in terms of temperatures above or below the glass transition temperature, Tg.25
The inevitable qualifier in this type of work is the fact that increasing rates of cooling
shift the Tg at higher temperatures and produce a less dense glass thus arguing that the
equilibrium glass condition lies below the experimentally accessible values.26
The next step of moving research in biomaterials towards the commercialization
stage encompasses largely the gelatin story. The protein, which for almost a century has
been produced on an industrial scale is the most frequently used structuring agent in
processed foodstuffs (e.g., confectioneries) and inactive drug additives (excipients) in the
pharmaceutical industry (e.g., the gelatin capsule).27,28
Reasons for the change in attitudes
include the need to circumvent diet and health problems or perceptions such as the BSE
scare, vegetarianism and religious dietary laws (e.g. Muslim and Hindu). Thus, there has
been an incentive to understand the molecular behavior of plant and seaweed
polysaccharides in a high-sugar environment since this class of biopolymers may provide
an alternative to gelatin.29
Work carried out in this regard identified and addressed three critical issues. Rubber
to glass transition was clearly identified with the glass transition temperature (Tg) being
defined as the conjunction of two molecular mechanisms. Further, the Tg measured by
differential scanning calorimetry (calorimetric Tg) remains unaltered in high solid
systems with low polysaccharide content (below 2.0%) and high levels (up to 90.0%) of
small-molecule co-solute. This suggests that the mobility of the co-solute (e.g., sugar) is
unaffected by the presence of the macromolecule.30
However, the mechanical profile of
the rubber-to-glass transition is strongly influenced by the polysaccharide particularly if it
is network forming.31
It is possible to represent the magnitude of this polysaccharide
76
contribution to mechanical properties by a ‘network Tg’, the greater the extent to which
this differs from the calorimetric Tg the larger the influence of the polymer on rheology.32
A conclusion has been made that the molecular nature of the glass transition is different
for DSC and rheology. While the calorimetric Tg is affected by the mobility of the sugar,
the rheological Tg is affected by the mobility of the biopolymer. This argument is valid
since both methodologies employed are fundamentally different and are sensitive to
different aspects of the glass transition. Lastly, the occurrence of considerable
disaggregation of the polysaccharide network has been suggested at high levels of co-
solute, resulting in a high degree of chain flexibility between cross-links and hence longer
texture.
Theoretical considerations adapted and utilized concepts from the sophisticated
synthetic polymer research but, increasingly, a stage is being reached at which it appears
possible to formulate an unexpected working hypothesis. Thus, quantification of the
effect of temperature and frequency on the structural properties of polysaccharide/co-
solute samples may document differences between their behaviour and data obtained for
the vitrification of amorphous synthetic materials. Consequently, there may be a break in
the parallel between the well established structural features of synthetics during
vitrification and high-solid biopolymers. The interplay of absorbed water and increasing
levels of co-solute could result in unexpected behavioral patterns, which are not in
accordance with the synthetic-polymer experience, and the present work evaluates this
hypothesis on the amylose/sugar system. In contrast to industrial polysaccharides and
gelatin, the morphology of the amylose network in a high sugar environment has not been
77
reported in the literature. An attempt has therefore been made to shed light on the
structural behaviour and gelation patterns of amylose in the presence of glucose syrup.
78
CHAPTER 2: EXPERIMENTAL SECTION
2.1 Materials
Amylose type III from potato essentially free of amylopectin was purchased from Sigma-
Aldrich, St. Louis, MO. Dilute amylose solutions in 1 M KOH were prepared to estimate
the intrinsic viscosity at 25°C, which was found to be 86 mL/g. The equation proposed
for amylose in pure aqueous solutions was used33,34
to yield a viscosity-average molecular
weight (Mv) of 4.8 x 105 dalton for the material of this investigation.
The glucose syrup (batch NX8472) used presently was a gift from Cerestar
(Mechelen, Belgium). It is prepared by controlled α-amylase hydrolysis of starch down
to glucose sequences linked by α-(1→4) glycosidic linkages.35
The functionality of the
product is determined by the dextrose equivalent (DE), which gives the content of
reducing end groups relative to glucose as 100. Thus DE of 25 corresponds to a (number
average) degree of polymerization (DP) of 4, and the DE of our sample is 42. The level
of solids is 83.0%, and the water content of glucose syrup was taken into account in the
calculation of the total level of solids for sample preparation. GPC analysis provided by
the manufacturer indicates the following relationship between DP and surface area (%) of
the glucose syrup spectrum:
DP1 17.54
DP2 12.99
DP3 10.55
DP4 8.79
DP5 7.29
DP6 5.28
DP7 4.78
DP8 4.21
79
DP9 3.19
DP10 1.98
>DP10 23.40
Total 100.0
2.2 Methods
Amylose samples were made by dissolving appropriate amounts of the polysaccharide in
1M KOH at 25°C using stirring for 30 min to achieve solvation. Concentrations ranged
from 2.0 to 4.5% and to each preparation 0.4% GDL was added to reduce the pH to 7.5 ±
0.1 at the end of the 90 min isothermal run. Gradual gelation of the amylose solution via
GDL addition was preferred to dropwise addition of HCl. This prevented problems
associated with loading on the measuring geometry of the rheometer arising from
instantaneous gel formation on addition of HCl. For mixtures, appropriate amounts of
glucose syrup were added to the amylose solutions to form a homogenous fluid using an
overhead stirrer and this was followed by addition of GDL.
Small deformation dynamic oscillation was used to obtain values of the storage
modulus on shear (G') which is the elastic component of the network, loss modulus (G";
viscous component) and dynamic viscosity (η*) of amylose samples with or without
glucose syrup. Variations with time and temperature were further assessed as a measure
of the ‘phase lag’ δ (tan δ = G" / G') of the relative liquid-like and solid-like structure of
the material.36 Measurements were performed with the Advanced Rheometrics Expansion
System (ARES), which is a controlled strain rheometer (TA Instruments, New Castle,
DE). ARES has an air-lubricated and essentially non-compliant force rebalance
transducer with the torque range being between 0.02 and 2,000 g cm.
80
Samples were loaded onto the parallel-plate geometry of the rheometer at 25°C and
their exposed edges covered with a silicone fluid from BDH (50 cS) to minimize water
loss. These were subjected to an isothermal run at the same temperature for 90 min at the
frequency of 1 rad/s and 0.1% strain, which was within the linear viscoelastic region. The
time sweep was followed by a frequency sweep between 0.1 and 100 rad/s. In addition,
high solid samples (e.g., 2% amylose plus 70% glucose syrup) were cooled to subzero
temperatures (down to – 60°C) at a rate of 1°C/min. This was followed by a heating scan
at 1°C/min to 25°C. In yet another experiment, materials were cooled to – 60°C but the
heating scans involved recording of isothermal frequency sweeps at fixed temperature
intervals of four degrees centigrade between 0.1 and 100 rad/s. In these, the applied strain
varied from 0.00072% in the glassy state to 1.0% in the flow region to accommodate the
huge changes in the measured stiffness of the sample. In doing so, parallel plate diameter
(from 40 to 5 mm) and measuring gap (3 + 1 mm) were also varied accordingly. Two
replicates were analyzed for each experimental preparation, with the melt-to-glass
transition being readily reproducible within a 3% error margin as a function of
temperature or timescale of measurement.
Modulated differential scanning calorimetry (MDSC) was carried out on 10 to 15
mg samples sealed hermetically in Alod-Al pans of Q1000 DSC (TA Instruments, New
Castle, DE). Cooling was controlled via a refrigerated unit (RCS) interfaced to the
calorimeter. The unit operates from – 90 to 550oC using a two-stage closed evaporative
system. Tzero calibration was performed by heating the cells without pans in the
temperature range of interest. Cell constant and temperature calibration was performed
with indium and MilliQ water at the heating rate of 1oC/min, and heat capacity was
81
calibrated using sapphire. For all scans, an empty pan was used as a reference and
nitrogen as a purge gas at the flow rate of 50 mL/min.37
For samples of low and intermediate levels of solids (e.g., 2% amylose or 4%
amylose plus 50% glucose syrup) the MDSC routine covered a temperature range of 0 to
80°C at the scan rate of 1°C/min. For high solid materials (e.g., 2% amylose plus 78%
glucose syrup), cooling scans were recorded at a controlled rate of 1°C/min to reach –
80°C. The pans were held at – 80°C for 30 min and scanned to 80°C with an underlying
heating rate of 1oC/min, period 40 s and amplitude ± 0.53°C described previously.
38
Measurements were carried out at least in triplicate yielding effectively overlapping
traces.
Scanning electron microscopy (SEM) was employed to observe morphological
features of the gels and to validate the fact that structural changes occur in the
biopolymer network formed with increasing levels of co-solute. For this, amylose
samples at various concentrations and their mixtures with a wide range of glucose-syrup
additions were first prepared and left to age at 25°C in order to form cohesive gels. These
were deep frozen rapidly by submerging into liquid nitrogen followed by freeze-drying
induced dehydration. Specimens were cut into thin slices before being mounted onto a
microscopy stub with a black double sided tape and coated with gold by means of a
sputter coater (JEOL JFC 1600 Fine Coater). Networks were observed using SEM, which
operated at an accelerating voltage of 15 kV (JEOL model JSM-5600LV, Tokyo,
Japan).39
82
CHAPTER 3: RESULTS AND DISCUSSION
3.1 Qualitative Aspects of the Effect of Increasing Levels of Co-Solute on the
Structural Properties of Amylose
As stated earlier, the effect of increasing levels of sugar on the structural properties of
various industrial polysaccharides as well as gelatin has been reported in detail in the
literature. In this part of the study, an attempt has therefore been made to illustrate and
contrast the properties of amylose/co-solute mixtures with those reported previously.
While industrial polysaccharides undergo readily a coil-to-helix transition, this study
suggests that amylose holds its structural characteristics unaltered at low and intermediate
levels of glucose syrup, hence enhancing the understanding of molecular behaviour in
high-solid systems. Neutralization of alkaline solutions of amylose with dropwise HCl
addition results in immediate non-reproducible gelation.4 Thus, utilization of HCl as the
acidulant made the study of the gelation properties of the amylose sample in question
rather difficult. Instead, a gradual reduction in pH was opted for using GDL, a slow-
release acidulant. This allowed proper sample loading on the measuring geometry of the
rheometer and well controlled induction periods leading to gel formation.
The development of three dimensional structures with time at 25°C and a frequency
of 1 rad/s is illustrated in Figure 3.1, where the logarithm of storage modulus (G') is
plotted against linear time over a 90-min period. The concentration (of amylose) spanned
a range between 2.0 to 4.5% and of note is the induction period, which is reduced with
increasing amounts of the polysaccharide in the system. Subsequently, a rapid
development of experimental traces followed by a steady but slow rise is observed
83
leading to a pseudo-equilibrium modulus at the end of the time of observation. Within
this concentration range, amylose preparations are smooth without exhibiting syneresis,
and the terminal G' values increase in excess of two orders of magnitude.
0.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8
3.2
3.6
0 20 40 60 80 100
Time (min)
Lo
g (
G'/
Pa)
Fig. 3.1. Storage modulus variation as a function of time of observation at 25°C for
amylose gels prepared at 2.0% (―), 2.5% (+), 3.0% (�), 3.5% (�), 4.0% (�) and 4.5%
(�) of the polysaccharide (frequency: 1 rad/s; strain: 0.1%).
Following the isothermal time sweep, supplementary experimentation unveiled the
effect of frequency of oscillation on the structural properties of single aqueous amylose
84
gels or of those in the presence of “low” additions of co-solute. Figure 3.2a makes the
case for a typical preparation of 4.0% amylose plus 30.0% glucose syrup subjected to
oscillatory frequencies of 0.1 to 100 rad/s. Traces show a solid-like behavior with G′ and
G" being approximately constant over the experimental frequency range, the loss tangent
(tan δ = G"/ G′) exhibiting a value of ≈ 0.045 throughout and the double logarithmic plot
of η* (dynamic viscosity) versus ω (angular frequency) having a gradient close to – 1.0
[Note: η* = (G' 2 + G" 2)1/2 / ω]. The mechanical spectrum was taken at an oscillatory
amplitude (strain) of 0.1%, which is well within the linear viscoelastic region. Unlike
conventional polysaccharides however, the amylose system exhibits an unusual strain
dependence. Indications of structural breakdown occur at strains as low as 1.0% for
amylose gels, an outcome which is not in accordance with experience from gels of
conventional coil-to-helix polysaccharides (e.g., alginates and κ-carrageenan).40,41
A plot of viscoelastic functions against oscillatory frequency obtained at 25oC for a
high-solid preparation comprising 2.0% amylose plus 70.0% glucose syrup is illustrated
in figure 3.2b. Clearly, the two profiles contrast strongly, with the high solids regime
exhibiting converging modulus traces, which lead to non-linear patterns of viscosity
development at the upper range of frequency. The pronounced frequency dependence of
the trace of loss modulus gives the superficial appearance of a “rubbery” structure that
ushers in the glass transition region at 100 rad/s (G" > G'). This characteristic
development of viscoelastic functions prompted us to carry out a systematic study of the
effect of co-solute content on amylose structure. In the 4.0% polysaccharide series,
homogeneous preparations reached a maximum glucose-syrup content of 50.0%, which
was extended to 78.0% in the 2.0% amylose series.
85
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
-1 0 1 2
Log (ω/rad s-1
)
Lo
g (
G',G
"/P
a;
η*
/Pa
s-1)
a
0.5
1.0
1.5
2.0
2.5
3.0
-1 0 1 2
Log (ω/rad s-1
)
Lo
g (
G',G
"/P
a;
η*
/Pa
s-1)
b
Fig. 3.2. Frequency variation of G' (�), G" (�) and η* (�) for a) 4.0% amylose plus
30.0% glucose syrup and b) 2.0% amylose plus 70% glucose syrup at 25°C (strain:
0.1%).
86
3.2 Amylose Diverging from the Paradigm of Coil-to-Helix Polysaccharides in a
High Solids Environment
Research on morphological characterization of polysaccharide networks in the
presence of increasing levels of sugar as the co-solute has been documented in the
literature. A dramatic collapse in the structural order of the gel networks at intermediate
levels of sugar is evident from the resulting plot of the normalized shear modulus of
mixtures as a function of sugar concentration. However, the trend for gelatin, the most
widely used structuring agent in the food industry is quite the opposite. The trace
obtained exhibits a rather radical reinforcement of the gelatin network at intermediate
levels of the co-solute. Akin to previous attempts, the normalized data for amylose gels as
a function of sugar concentration has been plotted alongside earlier results (with
permission from the author) in order to contrast the properties of amylose gels vis-à-vis
other gelling agents.
Figure 3.3 therefore reproduces normalized data of the storage modulus recorded
for various polysaccharide-sugar systems at ambient temperature. It constitutes an
attempt to summarize the variation in mechanical properties, as found in the literature and
the present work. Stronger structures of higher thermal stability are created as the level of
co-soulte is increased to 40% for gels of κ-carrageenan, agarose and deacylated gellan
gum.42
At such modest levels of sugar, conventional disorder to order transitions are
promoted resulting in a reinforcement of the rigidity of the polysaccharide network. As
can be observed, this trend of network reinforcement is also evident for the amylose
matrix.
87
-3
-2
-1
0
1
2
3
4
5
6
0 10 20 30 40 50 60 70 80 90 100
Sugar Concentration (%)
Lo
g (
No
rmal
ised
sh
ear
mo
du
lus)
coil-to-helix polysaccharide + sugar
sugar
amylose + sugar
Fig. 3.3. Variation of normalized storage modulus on shear (frequency of 100 rad s-1
) as a
function of sugar concentration for amylose/sugar (2% amylose (▲); 4% amylose (�)),
coil-to-helix polysaccharides (agarose/sugar, κ-carrageenan/sugar, deacylated
gellan/sugar), and single sugar preparations (original results and reference 7).
Interestingly, the nature of modulus development is entirely different at
intermediate levels of co-solute broadly defined between 40.0% and 65.0% solids. While
an abrupt drop is observed in the values of storage modulus for mixtures of sugar with the
industrial polysaccharides,7 amylose, maintains or accelerates the strengthening of the
network. This outcome yields a wide loop between the structural profiles of the two types
88
of systems. In the former, these observations can be rationalized by considering that
stabilization of stress-supporting polymer-polymer interactions require hydrogen bonding
from the surrounding (aqueous) environment. It appears, therefore, that the increasing
shortage of water molecules and their relatively stable hydrogen bonding with sugar
deprives gradually the polymeric associations of a hydration layer. X-ray diffraction and
other studies make a case that this layer is required for the thermodynamic stability of
intermolecular polysaccharide helices and the subsequent building of a stable
network.43,44
However such partial polymer disaggregation is not evident in the amylose
sequences, which instead maintain cohesion in the glucose-syrup environment. The
structural disparity between the two types of networks is further probed using differential
scanning calorimetry at a scan rate of 1°C/min as in the studies of mechanical
spectroscopy. Cooling runs of aqueous solutions of the industrial polysaccharides
produce well-defined exothermic events that vary in size, shape and temperature range.45
The area underneath these peaks can be used to obtain values of change in enthalpy (∆H)
of the coil-to-helix transition induced by cooling. Recently, it has been documented that
addition of co-solute at intermediate levels of solids (and beyond) results in broad
exotherms with a reduction in ∆H values, indicative of reduced order in the
polysaccharide network.46
None of the above is observed in the amylose thermograms. According to Figure
3.4, the temperature variation of heat flow for aqueous and intermediate-solid amylose
preparations exhibits a monotonic baseline drift down to 0°C. Clearly, the absence of a
coil-to-helix transition for amylose within the present experimental settings argues
89
against the necessity of hydrogen bonding with the aqueous environment as an essential
part of the process of structure development and stabilization. Further addition of co-
solute allows probing of molecular mechanisms involved in various relaxations of the
high solids preparation at the sub-zero regime.
0.5
0.6
0.7
0.8
0.9
-80 -60 -40 -20 0 20 40 60 80 100
Temperature (oC)
Hea
t fl
ow
(m
W)
2% am
4% am + 50% gs
80% gs
2% am + 78% gs
Fig. 3.4. Heat flow variation as a function of temperature for amylose, amylose/glucose
syrup and glucose syrup samples shown beside the individual traces (at a scan rate of 1°C
min-1
), am and gs stand for amylose and glucose syrup respectively.
90
As shown in Figure 3.4 for 2.0% amylose plus 78.0% glucose syrup or a single
glucose-syrup sample (80.0%), there is a well-defined change in heat flow (heat
capacity), which can be associated with the “endothermic” glass transition region upon
heating. The midpoint of this thermal event is readily detectable and is considered
presently to be the glass transition temperature, Tg, of the system. The DSC Tg remains an
empirical indicator of convenience and this has been discussed extensively in the
literature.47
DSC remains to be the most commonly employed methodology to obtain the
glass transition temperature of materials. It tracks changes in heat flow and the resultant
curve exhibits 3 points that could be reported- onset, midpoint of inflection or end-set of
the glass transition. The calorimetric Tg of amylose / glucose syrup mixture was found to
be – 42.1 ± 1.5°C whereas that of glucose syrup was - 45.8 ± 1.0°C. Both values are
comparable to those reported in the glass transition curve of the state diagram of sucrose
and glucose preparations at 80.0% solids.48
Referring back to Figure 3.3, the drop in the mechanical strength of industrial
polysaccharides at intermediate levels of sugar, primarily attributed to the transformation
from a highly aggregated structure to a network of reduced cross linking, is soon
followed by an equally spectacular increase in the values of storage modulus at the upper
range of co-solute (> 65.0% solids). At this higher range of co-solute, increased
flexibility of chain segments between intermolecular associations renders important the
entropic contribution to the network’s elasticity. The system is now seen as an
increasingly vitrified one where macromolecules form structures of reduced enthalpy of
cross-linking, shaping up a rubber-to-glass transition. The larger the influence of the
polysaccharide on the rheology the greater the acceleration of the mechanical
91
manifestation of Tg, as compared to the vitrification of single sugar systems at equal
solids content.7 Indeed, only at very high levels of solids (≥ 90%) the kinetics of sugar
vitrification match those of the polysaccharide-sugar mixtures owing to extreme removal
of water molecules from the polysaccharide domains so that formation of a three-
dimensional structure is abandoned.
On the other hand, the present mixture exhibits earlier signs of phase inversion from
a polysaccharide to a sugar dominated system. In Figure 3.3, formulations with 78.0%
glucose syrup and 2.0% amylose fall short in the magnitude of viscoelastic functions
expected from a polysaccharide continuous matrix. This result is further corroborated in
Figure 3.5, which reproduces the temperature profile of the same material at a controlled
scan rate of 1°C/min. At the highest experimental temperatures (> 0°C), G" dominates G'
thus unveiling part of the flow region. The rubbery plateau, i.e., flattened out and
dominant G' trace over the G" readings, which is characteristic of the temperature profile
of industrial polysaccharides in a high sugar environment,49
is not seen in amylose
samples. Instead, there is a direct passage from the terminal flow via the glass transition
region to the glassy state, with the traces of storage and loss modulus crossing over at
about – 35°C. This ushers in a hard solid-like response where the values of storage
modulus predominate and approach 1010 Pa.
92
2
3
4
5
6
7
8
9
10
11
-70 -50 -30 -10 10 30
Temperature (°C)
Lo
g (
G',
G"/
Pa)
G"
G'
Fig. 3.5. Cooling run of storage (�) and loss (�) modulus for 2.0% amylose in the
presence of 78.0% glucose syrup (frequency: 1 rad s-1
; scan rate: 1°C min-1
; strain 1.0%
to 0.00072%).
Further, tangible evidence of the disparity in the structural profile of our mixtures
with increasing levels of added glucose syrup is afforded using scanning electron
microscopy. Such micrographs for single amylose preparations and in mixture with co-
solute are reproduced in Figure 3.6.
93
A B
C D
E F Fig. 3.6. Scanning electron micrographs of A) 4.0% amylose, B) 4.0% amylose plus
15.0% glucose syrup, C) 4.0% amylose plus 30.0% glucose syrup, D) 4.0% amylose plus
50.0% glucose syrup, E) 2.0% amylose, and F) 2.0% amylose plus 78.0% glucose syrup.
Bar is 100 µm.
94
It appears that the aqueous amylose structures form dense fibrous assemblies of a
variable size depending on polymer concentration (4.0% and 2.0% in Figures 3.6A and
3.6E, respectively). Addition of co-solute changes gradually this densely packed
organization (Figures 3.6B and 3.6C). Thus the mixture of 4.0% amylose plus 50.0%
glucose syrup yields open, flattened and amorphous streams running through the
disorganized polysaccharide ridges in what appears to be a bicontinuous morphology in
Figure 3.6D. Further increments of glucose syrup (e.g., 78.0% solids with 2.0% amylose
in Figure 3.6F) result in phase inversion to a co-solute continuous matrix. This appears as
a smooth featureless background where noticeable remnants of the amylose molecular
organization are nonexistent.
3.3 Utilization of the Method of Reduced Variables to Quantify the Viscoelasticity of
Amylose-Sugar Mixtures During Vitrification
Discrepancies observed in the physical state and morphology of amylose / glucose syrup
samples when compared with those of intermediate / high-solid industrial
polysaccharides question the distinct behavioral patterns exhibited by the former. As in
the case of many synthetic polymers,50 the effect of temperature on viscoelastic features
of the glass dispersion produces a smooth and continuous pattern of development for
mixtures of co-solute and industrial polysaccharides.51
The following discussion
emphasizes the deviation of amylose-sugar gels from the mechanical response curves that
describe molecular relaxation mechanisms of other polysaccharides with similar
temperature dependence. For this purpose, results obtained for the formulation of 2.0%
amylose in the presence of 70.0% glucose syrup will be utilized.
95
2
3
4
5
6
7
8
9
10
11
-70 -50 -30 -10 10 30
Temperature (°C)
Lo
g (
G',
G"/
Pa)
0
1
2
3
4
5
Tan
δ
G"
G'
Tan δ
Fig. 3.7. Cooling run of storage (�) and loss (�) modulus and their ratio ‘tan δ’ (×) for
2.0% amylose in the presence of 70.0% glucose syrup (frequency: 1 rad s-1
; scan rate:
1°C min-1; strain 1.0% to 0.00072%).
Controlled cooling at 1°C/min results in a master curve of viscoelasticity illustrated
in figure 3.7, which is unseen in the vitrification of sugar / biopolymer mixtures. A clear
breakdown of thermorheological simplicity is encountered upon cooling with the data in
the glass transition region showing the obvious presence of different temperature
dependences. In the absence of a “single curve”, the first wave of structure formation
could be attributed to amylose (high molecular weight moiety) followed by the
vitrification of glucose syrup. Thus, two phenomena- polymer vitrification and rubber to
96
glass transition of the co-solute occur in succession and independently of each other. The
lowest range of temperatures where the glassy state is recorded allows mainly for
stretching and bending of chemical bonds.52 Furthermore, mechanical measurements of
the softening dispersion of industrial polysaccharide / sugar systems over a wide
temperature range return a single tan δ peak, an outcome that supports the concept of
thermorheological simplicity. In contrast, the present experimental window shows a main
peak at ≈ 4°C with another one at – 19°C (tan δ values of 4.26 and 1.94, respectively).
The low temperature “shoulder” eventually collapses to very low values as the material
exhibits the solid-like response of the glassy state (e.g., tan δ = 0.07 at – 60°C).
Rationalization of the viscoelastic ‘anomaly’ evident from the data in Figure 3.7 is
attempted next. The approach chosen in doing so may prove to be stimulating in guiding
research in other systems with an unexpected composite response during vitrification.
Working within the dynamic range of the mechanical oscillometer in our laboratory,
frequency sweeps were obtained at regular temperature intervals of four degrees
centigrade for each experimental wave of structure formation. Figures 3.8a and 3.8b
reproduce the outcome of this protocol that covers the isochronal temperature range in
Figure 3.7 but extends the effective timescale or frequency of observation from 0.1 to 100
rad/s. Storage and loss modulus data have relatively low values at high temperatures
(e.g., 0°C) whereas there is a substantial build up of small-deformation viscoelasticity
towards the low temperature end (e.g., - 40°C).
97
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Log (frequency / rad s-1
)
Log (
G' /
Pa)
a
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Log (frequency / rad s-1
)
Log (
G"
/ P
a)
b
Fig. 3.8. Frequency variation of a) G' and b) G" for 2.0% amylose plus 70.0% glucose
syrup. Bottom curve is taken at 0°C (�); other curves successively upwards – 8°C (�), -
16°C (�), - 20°C (�), - 24°C (�), - 28°C (�), - 32°C (�), - 36°C (�) and – 40°C (―).
Data at – 4°C, - 12°C, - 44°C, - 48°C, - 52°C, - 56°C and – 60°C have not been plotted to
avoid clutter.
98
2
3
4
5
6
7
8
9
10
-3 -1 1 3 5 7 9 11
Log (ωa T / rad s-1
)
Lo
g (
G' p
, G
" p /
Pa)
G'p
G"p
Fig. 3.9. Master curve of reduced shear moduli (G'p and G"p) for the sample of 2.0%
amylose plus 70.0% glucose syrup as a function of reduced frequency of oscillation (ωaT)
based on the frequency sweeps of Figure 3.8 (for plotting purposes the reference
temperature of - 4°C was used).
Characterization of molecular mechanisms focused on three distinct domains in
Figure 3.7 and the isothermal viscoelastic data at different temperatures in Figures 3.8a
and 3.8b were utilized to construct composite curves. Data were processed choosing
arbitrarily a point as the reference temperature (To) within each domain, and shifting the
remaining spectra along the log frequency axis until a uniform curve was obtained. Thus,
99
temperature essentially plays the role of changing the time or frequency scale and
assumes that a change in temperature within each domain shifts the time or frequency
scale of all molecular mechanisms by the same amount. The reduced composite curves
put together in the three temperature domains were further displaced horizontally for the
single purpose of the pictorial representation in Figure 3.9. This depicts a master curve of
viscoelasticity for an ultra-extensive range of oscillatory frequencies of about ten orders
of magnitude. There is also a combined increase in the values of reduced shear modulus
(G'p and G"p) of eight decades from 102 to ≈ 10
10 Pa.
The method of reduced variables employed in the construction of composite curves
that correspond to the distinct temperature ranges in Figure 3.7 yields three sets of shift
factors (aT) that can be treated with the popular concept of free volume in the form of the
Williams, Landel and Ferry (WLF) equation:53
log aT = -oo
oo
T - T + )/(f
)T - )(T(B/2.303f
fα
(1)
where, fo is the fractional increase in free volume at To, α
f is the thermal expansion
coefficient, and the value of B is set to be about one. The literature also defines the WLF
parameters o1C and o
2C as the ratios of B/2.303fo and f
o/αf, respectively. The WLF equation
holds for any temperature within the glass transition region including Tg, a result which leads
to the following relation:12
C1o = C
1
g C
2
g/(C
2
g + To - Tg)
C2o = C
2
g + To - Tg (2)
Thus the fractional free volume (fg) and the thermal expansion coefficient (αf) at Tg or
any other temperature within the glass transition region can be calculated.
100
An alternative approach to the theory of free volume is the predictions of the
reaction rate theory that can be formulated in terms of the modified Arrhenius equation:54
log aT = R303.2
aΕ (
oT
11−
Τ) (3)
where, R is the gas constant. This mathematical expression follows the progress in shift
factors by integrating two sets of temperature data. It returns a constant activation energy
(Ea), which argues that relaxation processes in the glassy state are heavily controlled by
specific chemical features.
Figure 3.10 reproduces the outcome of the application of this school of thought
(equations 1 to 3) to the viscoelastic functions of the amylose / glucose syrup mixture.
Three distinct patterns of structural relaxation have been featuring prominently, as
documented in the factor aT for the horizontal superposition of mechanical spectra in
Figures 3.8a and 3.8b. The high temperature fit has been taken by considering a reference
temperature of – 4.0°C and the WLF equation. This exponential function of the
reciprocal fractional free volume yields fg = 0.048, αf = 9.7 x 10
-4 deg and Tg1 = - 14.0°C.
Molecular dynamics are characterized by distinct kinetic rates in the second part of the
experimental temperature regime, which, however, follow well the predictions of the
combined WLF/free volume scheme; To = - 24.0°C, fg = 0.035, αf = 7.1 x 10-4 deg and
Tg2 = - 32.0°C.
It appears that in both cases, free volume is the molecular mechanism dictating
diffusional mobility in the system. It should be noted that the lower temperature
transition (Tg2) has a smaller value of fractional free volume (fg) and forms the more
dense glass. In conjunction with data in Figure 3.3, which argue against a molecular state
101
of mixing between amylose and glucose syrup sequences, the two glass transition
temperatures should be attributed to the morphology of a micro phase-separated material.
The implication is that the polymer vitrifies first followed by the degree of “order” frozen
into the glass structure formed by the co-solute.
-1
0
1
2
3
4
5
6
7
8
-55 -45 -35 -25 -15 -5 5
Temperature (°C)
Log a
T
amylose
WLF fit
Arrhenius fit
amylose T gglucose syrup T g
glucose syrup
WLF fit
Fig. 3.10. Temperature variation of the factor aT within the glass transition region of
amylose (�), the glass transition region of glucose syrup (�) and the glassy state (�) of
2.0% amylose plus 70.0% glucose syrup, with the solid lines reflecting the WLF and
modified Arrhenius fits of the shift factors throughout the vitrification regime (dashed
lines pinpoint the Tg predictions for amylose and glucose syrup).
102
At even lower temperatures, there is a clear change in the pattern of shift-factor
development towards a linear behaviour that cannot be followed by the WLF equation. In
here, the contribution of the mechanism of free volume is minimal and instead the
modified Arrhenius equation offers a viable alternative to describing molecular
dynamics. The latter yields a fixed value of the activation energy (650 kJ/mol) required
to overcome an energetic barrier for local rearrangements to occur from one state to
another. Such formalism assumes that the configurational entropy is essentially constant
within the glassy state.55
The present value of activation energy is higher than estimates
obtained by fitting the shift factor of amorphous synthetic polymers, which have been
found to be in the order of 200 kJ/mol, for example, in the vitrification of poly(ethylene
terephthalate).56 This discrepancy should be attributed to the high density of the bio-glass
owing to the stiffness of the polysaccharide chain and its ability to form compact double
helices.
103
CHAPTER 4: CONCLUSIONS & FUTURE TRENDS
Amylose gels are rather different than those obtained from other polysaccharides.
Industrial polysaccharides (e.g., agarose, deacylated gellan and κ-carrageenan), undergo
readily a coil-to-helix transition as a result of which a drastic drop in their mechanical
strength is observed in the presence of intermediate levels of co-solute. This has been
interpreted as a transformation from a highly enthalpic, aggregated structure to a network
of reduced cross-linking. Amylose, on the other hand holds its structural characteristics
unaltered at intermediate levels of glucose syrup following which is observed an early
phase inversion from polysaccharide to co-solute dominated system at levels of solids
above 70.0%. In the case of industrial polysaccharides however, kinetics of vitrification
are dictated upto levels of solids as high as 90.0% in the formulation. Additionally,
viscoelastic “anomalies” in the form of occurrence of two tan δ peaks in the passage from
the softening dispersion to the glassy state indicate a clear breakdown of
thermorheological simplicity. Further, mechanistic modeling argues for two distinct glass
transition temperatures in the mixture. It is therefore proposed that the amylose / glucose
syrup / water system does not reach a state of molecular mixing, with the morphological
features being those of a micro phase-separated material.
Gelatin replacement in foodstuffs has always been the key incentive to understand
the behaviour of polysaccharides in high sugar environments. With the use of gelatin
increasingly falling out of fashion with consumers and producers alike, polysaccharides
are invariably looked upon as alternatives. The behavioral response of gelatin in a high
solids environment has been shown to be rather different from that of polysaccharides,
104
with sugar promoting chain association rather than inhibiting it. This rather unique
behavioral response of gelatin in a high solids environment as compared to that of other
polysaccharides has been speculated as one of the reasons behind difficulties encountered
in its successful commercial replacement. Studying the behavioral pattern of amylose in
further detail based on evidence from this piece of work could provide a much needed
breakthrough as it appears to exhibit properties intermediary to those revealed by gelatin
and other polysaccharides. Findings would prove vital for applications in the
confectionery industry, and others such as flavour encapsulation, preservation/innovation
of glassy or rubbery foodstuffs, and preservation of bioactive molecules in glassy
carbohydrate matrices. In addition to the experimental techniques used in the current
study, X-ray diffraction studies could help ascertain the presence or absence and thereby
the role of crystallinity within amylose-glucose gels.
In conclusion, it is hoped that the working combination of phenomenological and
fundamental tools offers insights into the structural properties and phase topology of the
amylose / co-solute mixture. This can serve as a guide for similar work in other high-
solid preparations of contemporary interest, and to facilitate a direct link with
mechanistic models in low-solid composite gels where research endeavor and molecular
understanding is well advanced.
105
REFERENCES
1. Morris V.J. (1990). Starch gelation and retrogradation. Trends in Food Science
and Technology, 1, 2-6.
2. Miles M.J., Morris V.J., Orford P.D. and Ring S.G. (1985). The Roles of Amylose
and Amylopectin in the Gelation and Retrogradation of Starch. Carbohydrate
Research, 135, 271-281.
3. Miles M.J., Morris V.J. and Ring S.G. (1985). Gelation of Amylose.
Carbohydrate Research, 135, 257-269.
4. Doublier J-L., and Choplin L. (1989). A Rheological Description of Amylose
Gelation. Carbohydrate Research, 193, 215-226.
5. Ellis H.S. and Ring S.G. (1984). A Study of Some Factors Influencing Amylose
Gelation. Carbohydrate Polymers, 5, 201-213.
6. Anson K.J., Miles M.J., Morris V.J., Ring S.G. and Nave C. (1988). A Study of
Amylose Gelation Using a Synchroton X-Ray Source. Carbohydrate Polymers, 8,
45-53.
7. Kasapis, S.; Al-Marhoobi, I.M.; Deszczynski, M.; Mitchell, J.R.; Abeysekera, R.
(2003). Gelatin vs Polysaccharide in Mixture with Sugar. Biomacromolecules, 4,
1142-1149.
8. Slade L. and Levine H. (1990). Water relationships in starch transitions.
Carbohydrate Polymers, 21, 105-131.
9. Noel T.R., Ring S.G. and Whittam M.A. (1990). Glass transitions in low-moisture
foods. Trends in Food Science and Technology, 1, 62-67.
106
10. Kasapis S. and Al-Marhoobi I.M. (2005). Bridging the divide between the high-
and low-solid analyses in the Gelatin/k-Carrageenan mixture.
Biomacromolecules, 6, 14-23.
11. Kasapis S. (2005). Glass Transition Phenomena in Dehydrated Model Systems
and Foods: A Review. Drying Technology, 23, 731-757.
12. Ferry, J.D. (1980). In: Viscoelastic Properties of Polymers. New York: John
Wiley, USA.
13. Williams M.L., Landel R.F. and Ferry J.D. (1955). The temperature dependence
of relaxation mechanisms in amorphous polymers and other glass-forming liquids.
Journal of Amercian Chemical Society, 77, 3701-3707.
14. Whorlow R.W. (1992). In: Rheological Techniques, New York: Ellis Horwood,
2nd
Ed., 418-426.
15. Kasapis S. (2001). Critical assessment of the application of the wlf/free volume
theory to the structural properties of high solids systems: a review. International
Journal of Food Properties, 4(1), 59-79.
16. Hrma, P. (2008). Arrhenius model for high-temperature glass-viscosity with a
constant pre-exponential factor. Journal of Non-Crystalline Solids, 354, 1962-
1968.
17. Rahman, M.S.; Al-Marhubi, I.M.; Al-Mahrouqi, A. (2007). Measurement of glass
transition temperature by mechanical (DMTA), thermal (DSC and MDSC), water
diffusion and density methods: A comparison study. Chemical Physics Letter,
440, 372-377.
107
18. Singh, A.; Krogman, N.R.; Sethuraman, S.; Nair, L.S.; Sturgeon, J.L.; Brown,
P.W.; Laurencin, C.T.; Allcock, H.R. (2006). Effect of Side Group Chemistry on
the Properties of Biodegradable l-Alanine Cosubstituted Polyphosphazenes.
Biomacromolecules, 7(3), 914-918.
19. Li, Y.; Zhang, R.; Chen, H.; Zhang, J.; Suzuki, R.; Ohdaira, T.; Feldstein, M.M.;
Jean, Y.C. (2003). Depth Profile of Free Volume in a Mixture and Copolymers of
Poly(N-vinyl-pyrrolidone) and Poly(ethylene glycol) Studied by Positron
Annihilation Spectroscopy. Biomacromolecules, 4(6), 1856-1864.
20. Blachot, J.-F.; Chazeau, L.; Cavaille, J.-Y. (2002). Rheological behavior of
cellulose/monohydrate of N-methylmorpholine N-oxide solutions. Part 2. Glass
transition domain. Polymer, 43, 881-889.
21. Pawlus, S.; Paluch, M.; Kaminska, E.; Ngai, K.L. (2006). Anomalous narrowing
of the structural relaxation dispersion of Tris(dimethylsiloxy)phenylsilane at
elevated pressures. J. Phys. Chem. B, 110, 7678-7681.
22. Ngai, K.L.; Capaccioli, S.; Shinyashiki, N. (2008). The Protein “Glass” Transition
and the Role of the Solvent. J. Phys. Chem. B, 112, 3826-3832.
23. Mathew, A.P.; Dufresne, A. (2002). Plasticized Waxy Maize Starch: Effect of
Polyols and Relative Humidity on Material Properties. Biomacromolecules, 3,
1101-1108.
24. Barham, A.S.; Omar, A.M.E.; Roos, Y.H.; Kodnett, B.K. (2006).Evolution of
Phases and Habit during the Crystallization of Freeze-Dried Lactose/Salt
Mixtures in Humid Air. Crystal Growth & Design, 6, 1975-1982.
108
25. Angell, C.A. (2002). Liquid Fragility and the Glass Transition in Water and
Aqueous Solutions. Chem. Rev., 102, 2627-2650.
26. Allen, G. (1993). A history of the glassy state. In: The Glassy State in Foods
edited by J.M.V. Blanshard and P.J. Lillford, Nottingham University Press, 1-12.
27. Liu, L.S.; Liu, C.K.; Fishman, M.L.; Hicks, K.B. (2007). Composite Films from
Pectin and Fish Skin Gelatin or Soybean Flour Protein. Journal of Agricultural
and Food Chemistry, 55, 2349-2355.
28. Chiellini, E.; Cinelli, P.; Fernandes, E.G.; Kenawy, E.S.; Lazzeri, A. (2001).
Gelatin-Based Blends and Composites. Morphological and Thermal Mechanical
Characterization. Biomacromolecules, 2, 806-811.
29. Prestwich, G.D. (2008). Evaluating Drug Efficacy and Toxicology in Three
Dimensions: Using Synthetic Extracellular Matrices in Drug Discovery. Accounts
of Chemical Research, 41, 139-148.
30. Kumagai, H.; MacNaughtan, W.; Farhat, I.A.; Mitchell, J.R. (2002). The
influence of carrageenan on molecular mobility in low moisture amorphous
sugars. Carbohydrate Polymers, 48, 341-349.
31. Kasapis, S.; Mitchell, J.R. (2001). Definition of the rheological glass transition
temperature in association with the concept of iso-free-volume. International
Journal of Biological Macromolecules, 29, 315-321.
32. Kasapis, S. (2006). Building on the WLF/Free Volume Framework: Utilization of
the Coupling Model in the Relaxation Dynamics of the Gelatin/Cosolute System.
Biomacromolecules, 7, 1671-1678.
109
33. Kitamura, S.; Yoneda, S.; Kuge, T. (1984). Study on the retrogradation of starch.
I. Particle size and its distribution of amylose retrograded from aqueous solutions.
Carbohydrate Polymers, 4, 127-136.
34. Ring S.G.; l’Anson, K.J.; Morris, V.J. (1985). Static and dynamic light scattering
studies of amylose solutions. Macromolecules, 18(2), 182-188.
35. Tsoga, A.; Kasapis, S.; Richardson, R.K. (1999). The rubber-to-glass transition in
high sugar agarose systems. Biopolymers, 49, 267-275.
36. Malkin, A.Ya. and Isayev, A.I. (2006). In: Rheology – Concepts, Methods, &
Applications, ChemTec Publishing, Toronto.
37. Cui, S.W. (2005). In: Food Carbohydrates – Chemistry, Physical Properties, and
Applications, Taylor & Francis, Boca Raton.
38. Kontogiorgos, V.; Goff, H.D.; Kasapis, S. (2007). Effect of Aging and Ice
Structuring Proteins on the Morphology of Frozen Hydrated Gluten Networks.
Biomacromolecules, 8, 1293-1299.
39. Al-Amri, I.S.; Al-Adawi, K.M.; Al-Marhoobi, I.M.; Kasapis, S. (2005). Direct
imaging of the changing polysaccharide network at high levels of co-solute.
Carbohydrate Polymers, 61, 379-382.
40. Kong, H.J.; Kaigler, D.; Kim, K.; Mooney, D.J. (2007). Controlling Rigidity and
Degradation of Alginate Hydrogels via Molecular Weight Distribution.
Biomacromolecules, 5, 1720-1727.
41. MacArtain, P.; Jacquier, J.C.; Dawson, K.A. (2003). Physical characteristics of
calcium induced κ- carrageenan networks. Carbohydrate Polymers, 53, 395-400.
110
42. Watase M.; Nishinari K.; Williams P.A.; Phillips G.O. (1990). Agarose gels:
effect of sucrose, glucose, urea, and guanidine hydrochloride on the rheological
and thermal properties. Journal of Agricultural and Food Chemistry, 38, 1181-
1187.
43. Chandrasekaran R.; Radha A. (1995). Molecular architectures and functional
properties of gellan gum and related polysaccharides. Trends in Food Science and
Technology, 6, 143-148.
44. Janaswamy, S.; Chandrasekaran, R. (2006). Sodium ι-Carrageenan: A Paradigm
of Polymorphism and Pseudopolymorphism. Macromolecules, 39, 3345-3349.
45. Nitta, Y.; Kim, B.S.; Nishinari, K.; Shirakawa, M.; Yamatoya, K.; Oomoto, T.;
Asai, I. (2003). Synergistic Gel Formation of Xyloglucan/Gellan Mixtures as
Studied by Rheology, DSC, and Circular Dichroism. Biomacromolecules, 4,
1654-1660.
46. Al-Marhoobi, I.M.; Kasapis, S. (2005). Further evidence of the changing nature of
biopolymer networks in the presence of sugar. Carbohydrate Research, 340, 771-
774.
47. Liu, Y.; Bhandari, B.; Zhou, W. (2006). Glass Transition and Enthalpy Relaxation
of Amorphous Food Saccharides: A Review. Journal of Agricultural and Food
Chemistry, 54, 5701-5717.
48. Roos, Y.H. (1995). In: Phase Transitions in Foods. San Diego: Academic Press,
USA.
111
49. Ronan, S.; Alshuth, T.; Jerrams, S.; Murphy, N. (2007). Long-term stress
relaxation prediction for elastomers using the time–temperature superposition
method. Materials and Design, 28, 1513-1523.
50. Andreozzi ,L.; Castelvetro, M.; Faetti, M.; Giordano, M.; Zulli, F. (2006).
Rheological and Thermal Properties of Narrow Distribution Poly(ethyl acrylate)s.
Macromolecules, 39, 1880-1889.
51. Haque, M.K.; Kawai, K.; Suzuki, T. (2006). Glass transition and enthalpy
relaxation of amorphous lactose glass. Carbohydrate Research, 341, 1884-1889.
52. Robertson, C.G.; Palade, L.I. (2006). Unified application of the coupling model to
segmental, Rouse, and terminal dynamics of entangled polymers. Journal of Non-
Crystalline Solids, 352, 342-348.
53. Ferry, J.D. (1991). Some reflections on the early development of polymer
dynamics: viscoelasticity, dielectric dispersion, and self-diffusion.
Macromolecules, 24, 5237-5245.
54. Mano, J.F.; Viana, J.C. (2006). Stress–strain experiments as a mechanical
spectroscopic technique to characterise the glass transition dynamics in
poly(ethylene terephthalate). Polymer Testing, 25, 953-960.
55. Alves, N.M.; Mano, J.F.; Gomez Ribelles, J.L; Gomez Tejedor, J.A. (2004).
Departure from the Vogel behaviour in the glass transition—thermally stimulated
recovery, creep and dynamic mechanical analysis studies. Polymer, 45, 1007-
1017.
56. Alves, N.M.; Mano, J.F.; Gomez Ribelles, J.L. (2002). Molecular mobility in
polymers studied with thermally stimulated recovery. II. Study of the glass
112
transition of a semicrystalline PET and comparison with DSC and DMA results.
Polymer, 43, 3627-3633.