University of Kalmar
School of Pure and Applied Natural Sciences
Interactions between hydrophobically
modified starch and egg yolk proteins in
solution and at oil/water interfaces
Emma Magnusson
Degree project work in chemistry
Level: D
Degree project works made at the University of Kalmar, School of Pure and Applied Natural Sciences, can be ordered from: www.hik.se/student
or
University of Kalmar
School of Pure and Applied Natural Sciences SE-391 82 KALMAR
SWEDEN
Phone + 46 480-44 62 00 Fax + 46 480-44 73 05 e-mail: info@nv.hik.se
1
Interactions between hydrophobically modified starch and egg yolk
proteins in solution and at oil/water interfaces
Emma Magnusson
Chemistry 240 hp
University of Kalmar, School of Pure and Applied Natural Sciences. Degree project work, 30 hp
Supervisors: Lars Nilsson, Ph. D. Food Technology, Faculty of Engineering University of Lund
SE- 221 00 LUND Sweden
Päivi Jokela, Ph. D. School of Communication and Design
University of Kalmar SE-391 82 KALMAR Sweden
Examiner: Kjell Edman, Ph. D. School of Pure and Applied Natural Sciences
University of Kalmar SE-391 82 KALMAR Sweden
Abstract
A common modification of starch is esterfication with anhydrous octenyl succinic acid (OSA). The modification makes the polymer surface active and it also incorporates a carboxyl group to the starch, which can be negatively charged. The characteristics of OSA starch make it interesting for usage in combination with egg yolk proteins in food emulsions. It is not only the individual ingredients that affect the product; interactions between
ingredients and ingredient-dispersion medium have a great impact on factors such as structure and stability. Knowledge about how the interactions affect emulsion properties would make it possible to predict the behavior of an emulsion, which would be a great advantage in the formulation of food emulsions. Therefore, this is a subject of interest.
The purpose of this master thesis was to further investigate the interactions between OSA starch and α – β-livetin in solutions and in emulsions. First, the charges of the macromolecules were studied by titration. Interactions in solution were then analyzed through turbidity and solubility measurements. The adsorption of OSA starch onto livetin and the interfacial rheology were also studied. Finally, an emulsion stability experiment was made.
Strong interactions between the two macromolecules were observed in solutions at pH 4.0. This was probably due to hydrophobic interaction; however it could also be explained by electrostatic interaction. In the emulsions the adsorption of starch onto livetin was highest at pH 4.5, and then decreased with increasing pH values. The absence of OSA starch adsorption at pH 4.0, despite the strong interaction in solution, could be explained by complex formation immediately in solution. Less starch would then be able to reach the interface and adsorb. In the interfacial rheology experiments, an indication of decreased complex dilational modulus of the interfacial layer, caused by OSA starch addition was seen at low pH values. This could be due to aggregation of the proteins and formation of an uneven interfacial layer. OSA starch would then be able to adsorb and disturb the elasticity. Some differences in the stability of an emulsion only containing livetin, and an emulsion with both livetin and OSA starch could be observed. However, more investigations are needed to be made to understand the underlying mechanisms.
2
SVENSK SAMMANFATTNING
Hydrofobmodifierad stärkelse, såsom oktenyl bärnstenssyreanhydrid (OSA) stärkelse används idag i ett flertal livsmedel. Modifieringen gör stärkelsen amfifil och den kan därmed användas för formulering och stabilisering av emulsioner. OSA stärkelse har visats kunna ge en ökad värme- och frysstabilitet för livsmedel. OSA stärkelsens egenskaper gör den intressant för användning i livsmedelsemulsioner stabiliserade med ägguleproteiner, såsom majonnäs och såser. En produkts egenskaper avgörs inte bara av de enskilda
ingredienserna, utan även av interaktionerna mellan dem. Syftet med detta arbete är därför att studera interaktioner mellan OSA stärkelse och en vattenlöslig proteinfraktion av äggulan (α-β-livetin), och se hur dessa påverkar emulsionsegenskaper.
För att kunna utröna vilka interaktioner som skulle kunna förekomma mellan OSA stärkelse och α-β-livetin studerades laddningen på makromolekylerna vid olika pH utifrån titrerkurvor. Interaktionerna i lösning utforskades därefter med hjälp av turbiditets- och löslighetsmätningar. Därefter studerades även interaktionerna i emulsioner; adsorptionen av OSA stärkelse på en gränsyta täckt av α-β-livetin undersöktes och ytreologiska mätningar genomfördes. Slutligen utfördes en emulsionsstabilitetsstudie.
En stark indikation på komplexbildning mellan OSA stärkelse och α-β-livetin sågs vid pH 4.0. Denna berodde troligtvis på hydrofoba interaktioner men även elektrostatiska
interaktioner är en möjlig förklaring. I emulsioner sågs den största adsorptionen av OSA stärkelse på α-β-livetin vid pH 4.5, för att sedan minska med ökande pH värden.
Avsaknaden av adsorption av stärkelsen på α-β-livetin vid pH 4.0, trots stark interaktion i lösning, kan bero på att komplex bildades direkt i lösning och mindre stärkelse därmed nådde gränsytan och kunde adsorbera. I de ytreologiska mätningarna sågs en minskning av ytskiktets elasticitet vid stärkelsetillsats till en emulsionsdroppe stabiliserad med α-β-livetin vid låga pH värden. Detta kan ha berott på ett ojämnt adsorberat lager av protein och bildning av små luckor i ytskiktet. OSA stärkelse skulle därmed ha möjlighet att adsorbera och påverka elasticiteten. I stabilitetsstudien sågs små skillnader mellan emulsioner endast innehållande α–β-livetin och emulsioner med både OSA stärkelse och α–β-livetin. För att förstå de underliggande mekanismerna krävs dock fler studier.
Förståelse för hur interaktioner påverkar emulsionsegenskaper skulle ge stora fördelar för formuleringen av emulsioner.
3
PREFACE
This 20 week long master thesis was performed as a part of the Master Degree program in Nutrition and Food Science at the University of Kalmar. The practical work was done at the Department of Food Technology at the Faculty of Engineering at Lund University. I would like to thank my supervisor at Lund University, Lars Nilsson, for all help and support and Päivi Jokela at the University of Kalmar for her input in the writing of this report.
I would also like to thank all other employees at the Department of Food Technology for making me feel welcome at the department, and a special thank you, to those who have helped me with the equipment and so on.
At last, I would like to thank my friends and family for listening and being there for me.
4
CONTENTS
ABBRIVATIONS AND SYMBOLS ... 6
1. INTRODUCTION ... 7 1.1 Aim ... 7 1.2. Modified starch ... 8 1.2.1. OSA-Starch ... 8 1.3. Egg-yolk proteins ... 9 1.3.1. α – β-livetin ... 10
1.4. Interactions between proteins and polysaccharides ... 10
1.5. Emulsions ... 11
1.5.1. Emulsion stability ... 12
1.5.2. Proteins and emulsions ... 13
1.5.3. Polysaccharides and emulsions ... 14
1.5.4. Stabilization or destabilization? ... 15
1.5.5. Adsorption and desorption ... 16
1.6. Interfacial rheology... 16
2. MATERIALS AND METHODS ... 18
2.1. Dynamic Drop Tensiometer ... 18
2.2. Coulter light scattering apparatus ... 19
2.3. Zetasizer ... 19 2.4. Chemicals ... 21 2.5. Preparation of solutions ... 21 2.6. Titration curves ... 21 2.7. Interactions in solution ... 22 2.7.1 Transmission profiles ... 22 2.7.2 Solubility ... 22 2.8. Interactions in emulsions ... 22 2.8.1. Adsorption isotherms ... 22
2.8.2. Analyzing α-β-livetin and OSA-starch solutions with the drop tensiometer ... 24
2.8.3. Emulsion stability ... 24
3. RESULTS ... 25
3.1. Titration curves ... 25
5 3.2.1 Transmission profiles ... 27 3.2.2 Solubility ... 30 3.3. Interactions in emulsions ... 30 3.3.1. Adsorption isotherms ... 30 3.4. Interfacial rheology... 33
3.4.1. Analyzing α-β-livetin and OSA-starch solutions with the drop tenisometer ... 33
3.5. Emulsion stability ... 34
4. DISKUSSION ... 36
4.1. The charge of the macromolecules ... 36
4.1. Interactions in solution ... 36
4.2. Interactions in emulsions ... 39
5. CONCLUSIONS ... 43
REFERENCES ... 44
6
ABBRIVATIONS AND SYMBOLS
A Emulsion interfacial area (m2)
a Air
BET Brunauer, Emmet and Teller
Cads Adsorbed amount (mg/m3)
Ctot Total concentration in an emulsion (g/L)
C0 Initial bulk concentration (g/L)
DS Degree of substitution, defined as average number of esterfied hydroxyl groups per monosaccharide unit
d32 Area-weighted particle diameter (µm)
E Dilational modulus/complex modulus (mN/m)
E’ Storage modulus (mN/m)
E” Loss modulus (mN/m)
Ed Dilational elasticity (mN/m)
OSA Octenyl succinic anhydride
o Oil
pI Iso-electric point of a protein
SD Standard deviation
µ Electrophoretic mobility
w Water
ε Permittivity (C² N-1 m-2)
Г Surface load (mg/m2)
η Dynamic viscosity (Pa s)
ηd Interfacial dilational viscosity (Pa s)
γ Surface/interfacial tension (mN/m)
ω Angular frequency (Hz)
ζ Zeta-potential (mV)
7
1. INTRODUCTION
A common modification of starch is esterfication with anhydrous octenyl succinic acid (OSA). The modification makes the polymer surface active; thus it can affect the formation and stabilization of dispersed food systems, such as emulsions. (Nilsson and Bergenståhl 2006)
The properties of the OSA starch make it interesting for used in egg yolk protein stabilized food emulsions, such as mayonnaise. The interactions between different components in a product have a great impact on factors such as texture and stability (De Kruif and Turnier 2001) and it is thereby interesting to study the interactions between these two
macromolecules*.
The modification of the starch also incorporates a carboxyl group onto the starch, which can be negatively charged (Nilsson and Bergenståhl 20062). Previously, Nilsson et al. (unpublished material) made an experiment where they studied the adsorption of OSA starch onto an oil-water interface covered with egg yolk proteins (α – β-livetin). They expected a high adsorption of OSA starch onto livetin at low pH values, because of the opposite charges of protein and polysaccharide, and a low adsorption at high pH values because of the same charges of the macromolecules. However, their experiment showed exactly opposite results; the protein adsorbed more at higher pH values and the adsorption did thereby not seem to depend on electrostatic interactions. Their results also indicated that the OSA starch could displace the proteins at the interface.
1.1 Aim
The purpose of this master thesis is to further investigate the interactions between OSA starch and α-β-livetin, both in solutions and in emulsions. The aim is to get better knowledge of what kind of interactions that exist between the macromolecules and how these interactions affect emulsions.
Methods used were; to study titration curves, in order to determine the charge of the macromolecules at different pH values; turbidity measurements, to study the aggregation between the macromolecules; and solubility experiments, to investigate how the
macromolecules affect the solubility of each other. Adsorption isotherms of OSA
8 starch in livetin stabilized oil-in-water emulsions were also studied, to see how the starch adsorbs onto the livetin layer. In addition, the change of the interfacial rheology after addition of starch to an air drop covered with livetin were studied, to observe possible desorption of the proteins. Finally, a stability study was made to see how the addition of starch affected flocculation and coalescence of oil-in-water emulsions.
1.2. Modified starch
Starch is the polysaccharide used for storage of energy in plants and it also provides humans worldwide with a great part of their daily calorie intake. Starch consists of two different polymers, amylose and amylopectin. Amylose is a mainly linear polymer consisting of units of α-D-glucopyranose with (14) glycosidic links and a few branches attached in (16) position (Takeda et al. 1990). Amylopectin has the same structure as amylose except that the polymer is more branched, and hence it is much larger (Nilsson and Bergenståhl 2006). Starch, both in its native and modified forms, has a number of applications in foods; such as thickening, stabilizing and texturing (BeMiller and Whistler 1996).
Starches are widely modified to achieve increased process and storage stability in foods. Starches can also be used as emulsion stabilizers by adding hydrophobic side chains to the starch molecule. This renders the normally hydrophilic starch molecule an amphiphilic character. (Tesch et al. 2002)
1.2.1. OSA-Starch
The modification of starch through esterfication with dicarboxylic acids was patented in 1953 by Caldwell and Wurzburg (Caldwell and Wurzburg 1953). A common chemical modification of starch in order to achieve an amphiphilic molecule is esterfication of the starch and anhydrous octenyl succinic acid under alkaline conditions (Trubiano 1986). Octenyl succinic anhydride starch, also called OSA-starch is an approved food additive in EU and goes under the E-number 1450 (Tesch et al. 2002). The chemical structure of OSA starch is shown in Figure 1.
The substitution with OSA can occur at carbon 2, 3 and 6 in the glucose molecule (Nilsson and Bergenståhl 20061). According to Shogren et al. (2000), OSA groups are probably mostly present in the amorphous parts of the amylopectin molecule, in the interior of the starch; but they also exist on the exterior of the granule.
9 O H O O H O O O O H OH O O H O O O O
Figure 1. The chemical structure of octenyl succinic anhydride (OSA) starch.
A degree of substitution* ~ 0.02 is typical for commercial samples (Shogren et al. 2000).
Moreover, Viswanathan (2009) showed that a higher degree of substitution does not necessarily improve the emulsification activity.
OSA starches have different properties dependent on the botanic origin of the starch and the DS. However, the main outcomes of the substitution is increased viscosity and decreased gelatinization temperature, and as mentioned earlier, the ability to stabilize emulsions (Bao et al. 2003). OSA starch may give heat- (He et al. 2008) and freeze-thaw stability (Song et al. 2006), and may also be used as a partial replacement for fat, as it may give a sense of fattiness (BeMiller and Whistler 1996). Moreover, Heacock et al. (2004) found that the esterfication of the starch decreased the extent of starch digestion and that OSA starch may be used as a Functional Fiber. Beverage emulsions (Ashurst 2005), emulsion concentrates of flavors and heavy salad dressings are examples of uses for the starch (Stephen et al. 2006).
The OSA-starch used in this master thesis was of waxy barley origin, consisting of 7% amylose and 93% amylopectin. Nilsson and Bergenståhl (2006) determined the molar degree of substitution to 0.0213 for the sample.
1.3. Egg-yolk proteins
Because of their extraordinary characteristics, egg yolk proteins are used in a great variety of food stuff. The egg yolk proteins provide high nutritional value at a relatively low price and they can give desired texture modifications. The egg yolk proteins are also surface active and they can be used to form stable foams and emulsions. (Bell et al. 2002)
10 Egg yolk consists of several proteins and it also contains other surface active components such as phospholipids and cholesterol (Nilsson et al. 2007). This makes total hen egg yolk a very complex system and also hard to work with. Therefore, a water soluble fraction, α-β-livetin, from the egg yolk plasma has been used as a model system in this master thesis.
1.3.1. α-β-livetin
The α-β-livetin fraction of the egg yolk plasma can be achieved through stepwise precipitation with NaCl and centrifugation (McBee and Cotterhill 1979). Nilsson et al. (2006) analyzed the fraction using two-dimensional polyacrylamide gel electrophoresis and mass spectrometry. They found that the fraction dominantly consist of five protein species; ovotransferrin, IgG (heavy chain), serum albumin, YGP42 and YGP40. The molecular masses of the proteins were determined to be in the range from 37 to 80 kDa and the pI from five to eight, as shown in Table 1.
Table 1. Five dominant protein species in the livetin fraction as determined by 2D SDS-PAGE and subsequent mass spectrometry by Nilsson et al. (2006)
1.4. Interactions between proteins and polysaccharides
Macroscopic properties of food products such as flow, stability, texture and mouth feel do not only depend on the properties of the individual ingredients, but also on the interactions between them (de Kruif and Turnier 2001).
When mixing proteins and polysaccharides in solution, three main possibilities of behavior exist. In mixtures of very dilute solutions, the protein and polysaccharide are co-soluble. The mixing entropy dominates and the system is stable. At higher macromolecule concentrations in the solution, the system might become unstable. (De Kruif and Turnier 2001) If the protein and polysaccharide have the same charge they repel each other, thermodynamic incompatibility takes place and the macromolecules tend to segregate into
Protein Observed molar mass (kDa) Observed pI Ovotransferrin (conalbumin) 80 6,5 - 7
IgG, heavy chain 65 -70 6,5 - 8
serum albumin (α -livetin) 65 5 - 5,7
YGP42 40 5,3 - 5,8
11 two phases (Neirynck et al. 2007). The interaction can also be associative, which means that the macromolecules attract one another and form complexes (de Kruif and Turnier 2001), for example due to hydrophobic and ionic interactions. These complexes can either be soluble or insoluble; in the latter case, aggregative phase separation may occur.
(Neirynck et al. 2007) A B C
Figure 2. Main possibilities of behavior for the mixing of polysaccharide and protein. A; segregation, B; co-solubility, C; association.
Combinations of proteins and polysaccharides are often used as ingredients in food emulsions, and the interactions between the biopolymers may have a significant impact on the emulsion characteristics (Lael-Calderon et al. 2007). Therefore, it is interesting to study the interactions between these macromolecules in emulsions.
1.5. Emulsions
An emulsion is defined as a heterogeneous system with two immiscible, or poorly miscible fluids (Shaw 1992). One of the fluids is dispersed into the other as small droplets. The droplets are of colloidal size (i.e. below 100 µm), and typically between 0,1 µm and 10 µm in diameter in the usual type of emulsion. (Shaw 1992, Dimitrova et al. 2004)
Emulsions find a variety of applications in for example food, cosmetics and
pharmaceutical preparations. Emulsions can give enhanced control over factors such as rheology, taste etc. It is also possible to dilute an expensive or concentrated ingredient with an inexpensive but immiscible one by making an emulsion. (Shaw 1992)
12 Almost all emulsions consist of an aqueous phase and an oil phase (in the widest meaning of the term). Oil-in-water (o/w) emulsions consist of small droplets of oil dispersed in the water phase, i.e. the continuous phase is aqueous and the dispersed phase is oil, whereas water-in-oil (w/o) emulsions are water droplets dispersed in an oil phase. (Shaw 1992) Dressings and soups are examples of oil-in-water emulsions whereas butter and margarine are w/o emulsions (Walstra 1996). During this thesis, o/w emulsions are the ones to be studied.
An emulsion is usually formed by adding intense mechanical energy to the phases. This can for example be made by high pressure homogenization. An emulsion formed with to pure liquids is not stable and will soon go through phase separation. For an emulsion to be stable, adsorption of stabilizing molecules (or solid particles (Hunter et al. 2004)) at the droplets must occur simultaneously as the formation of interfacial area. (Nilsson 2007)
1.5.1. Emulsion stability
All emulsions are thermodynamically unstable systems. This means that their decay results in a decrease of the free energy and they will eventually break, i.e. the two phases will separate (Hunter et al. 2004).
Emulsion stability is a very complex phenomenon. The stability is affected by many factors, such as the droplet size distribution (Tesch et al. 2002), the viscosity of the continuous phase and interactions between particles and between particles and the dispersion medium (Shaw 1992). Ionic strength and pH also affect the emulsion stability, for example by altering electrostatic interactions. Because of the thermodynamic instability of emulsions, the droplet size tends to increase in order to reduce the total free energy (Capek 2004). Many forms of instabilities such as creaming, flocculation and coalescence may occur. Creaming is caused by the difference in density of the oil and water phases and the gravity thereby induces the rising of the oil in the o/w emulsion. Flocculation is the interaction between droplets which leads to the formation of clusters without fusion of the droplets. Coalescence is the irreversible fusion of emulsion droplets that leads to the formation of one large drop and if the coalescence continues the emulsion eventually breaks. (Shaw 1992, Hunter et al. 2008)
13 A.
B.
C.
Figure 3. Different forms of destabilisation of emulsions. A; creaming, B; flocculation, C; coalescence
Many emulsions, however, remain in a dispersed form for a long time because of kinetic stability (McClements 2004). The covering of the interface with emulsifying agents makes up a kinetic barrier and prevents phase separation (Dimitrova et al. 2004). The emulsifying agents are generally molecular surfactants, polymers and proteins as well as finely divided solids (Hunter et al. 2004). The surfactants, polymers and proteins have both hydrophobic and hydrophilic parts and they are thereby drawn to the oil/water interface. The adsorption of molecules at the interface causes a lowering of the interfacial tension, and thereby the interfacial energy (Shaw 1992). Low molecular weight surfactants in high concentration lower the interfacial tension more than proteins, because of the more efficient packing of the small molecules at the interface (Dimitrova et al. 2004). Proteins and polymers do, however, stabilize emulsions through electrostatic and steric stabilization (Nilsson et al. 2006). When large molecules cover the surface of the droplets, steric repulsions between droplets may occur, and coalescence is thereby prevented. This steric stabilization depends on the size and the solubility of the adsorbed molecule in the continuous phase. (Nilsson and Bergenståhl 20061) Electrostatic stabilization occurs in o/w emulsions when similarly charged electric double layers overlap, which will lead to repulsion between the dispersed particles (Shaw 1992).
1.5.2. Proteins and emulsions
Proteins are the natural choice of emulsifier to stabilize o/w emulsions, because of their high surface activity and superior ability to resist coalescence. They do not lower the
14 interfacial tension significantly but they provide considerable electrostatic and steric repulsion. Because of their poor solubility in the oil phase, proteins cannot be used for w/o emulsions. (Walstra1996)
When proteins have adsorbed at the interface, they gradually start to unfold. They do this because of the presence of a hydrophobic fluid phase, oil, which makes it possible for the protein to turn its hydrophobic parts towards the oil and thereby minimize its
configurational free energy. The unfolding of the proteins leads to uncovering of different segments and thereby facilitates lateral interactions between the adsorbed molecules. Different interactions may take place such as ionic and hydrophobic interactions as well as disulfide bridging and hydrogen binding. Unique for proteins are that a viscoelastic network, with nanometer thickness is formed and that this network protects the droplets against coalescence. Lateral interactions like these do not exist, or are at least considered minimal, between the low molecular weight surfactants. (Dimitrova et al. 2004)
Proteins are generally less efficient as emulsion stabilizers under conditions of low pH, high ionic strength and high temperature since these conditions lower the solubility of the proteins (Nilsson and Bergenståhl 2007). Egg yolk, however, contains protein species with a broad pI range (<4 - >9) and it is therefore able to stabilize emulsions at lower pH values (Nilsson et al. 2007).
Nilsson et al. (2006) examined the possibility to form stable emulsions with the α-β-livetin fraction from the egg yolk plasma. They found that the proteins worked as emulsifiers and reached a surface load (se section 1.5.6.) of 1,5 mg/m2 at pH 7, as can be expected for monolayer formation of a protein (Walstra 1996).
1.5.3. Polysaccharides and emulsions
Polysaccharides mostly act as stabilizers in emulsions by enhancing the viscosity of the aqueous phase (Lael-Calderon et al. 2007). A higher viscosity of the continuous phase conteracts to creaming by slowing down the rising of the oil droplets in the emulsion (Walstra 1996). OSA-starch can act as a viscosity enhancer, but its thickening capacity is quite limited compared to the capacity of other macromolecules used as thickeners (Erni et al. 2007). However, the combination of emulsifying and thickening properties of OSA starch can give a cost reduction of the final product because a lower concentration of stabilizing agents is needed (Tesch et al. 2002).
15 Hydrophobically modified polysaccharides, such as OSA starch, are surface active. The driving force for the adsorption of the modified polysaccharide at the oil droplets is mainly hydrophobic interaction (Nilsson and Bergenståhl 20061). When adsorbed, the polysaccharide induces steric stabilization due to its high molar mass and branched polymer structure (Nilsson et al. 20062). Tesch et al. (2002)found it possible to form stable emulsion with OSA starch concentrations as low as 0,08 wt% and claimed that steric stabilization is the main stabilizing mechanism of OSA starch, as the emulsification was independent of pH value and ionic strength. Nilsson and Bergenståhl (20062) observed extremely high surface loads (up to 40 mg/m2) of OSA starch when allowed to adsorb at a
positively charged interface. This indicates that OSA starch could be suitable for encapsulation.
1.5.4. Stabilization or destabilization?
Although emulsifiers are necessary to provide stable emulsions, they may also contribute to destabilization of the emulsions. The decrease in interfacial tension, which stabilizes the emulsion by lowering the interfacial free energy, also makes it easier to disrupt droplets. (Walstra 1996)
In the case of polymers in emulsions, non-adsorbed polymers may cause depletion flocculation. The large polymers cannot come too close to an interface and the polymer concentration near the interface is thereby lower, a depletion zone is formed. The osmotic pressure will give rise to a movement of water from the depletion zones and into the more concentrated polymer solution. If the layers with low concentration of polymer come near each other, they overlap, and this causes a lowering of the free energy. (Fennema 1996, Somasundaran 2006) The polymers may also destabilize emulsions by forming bridges between particles. This happens when the covering of the interface by polymer is insufficient whereby the polymer may adsorb onto more than one droplet. (De Kruif and Turnier 2001)
Addition of emulsifying components may either stabilize or destabilize the emulsion and it seems that the effects of different surfactants cannot be generalized; they depend largely on specific aspects of interactions between different emulsifying agents and the properties of the adsorbed layer at the interface. (Derkatch et al. 2007)
16
1.5.5. Adsorption and desorption
The emulsifying agents are transported to the interface by diffusion or convection, or both (Benjamins 2000). The rate of the adsorption is important because it affects the amount of emulsifying agents to be adsorbed and the properties of the adsorbed layer. For example, a slow adsorption gives a greater unfolding of a protein, the protein has time and space to unfold, whereas a faster adsorption gives a closer packing and less space to unfold. (Miller et al. 2005)
The adsorption of an emulsifying agent can be expressed as an adsorption isotherm with the surface load (Г), which is the adsorbed amount of emulsifying agents, as a function of the equilibrium bulk concentration of the agent (Nilsson and Bergenståhl 2006). Single layer adsorption generally gives a surface load of 1-3 mg m-2(Walstra 1996) and is expressed as a Langmuir adsorption isotherm (Buckton 1995). Multilayer formation can give very high surface loads and are more often expressed as BET isotherms (Buckton 1995).
Protein adsorption is generally seen as an irreversible process. However, adsorbed protein species may be replaced by other proteins after the adsorption. This is dependent of the affinity of the proteins to the interface; proteins with weak affinity can be replaced by proteins that are more favorable energetically. (Nilsson et al. 2006) Desorption of the proteins is, however, less likely after the unfolding of proteins (Van der Veen et al. 2006). During this thesis, possible desorption of proteins will be observed by studying the interfacial rheology.
1.6. Interfacial rheology
The adsorption of emulsifying agents at an interface gives the interface rheological properties (Walstra 1996). Interfacial rheology is defined as the study of the deformation and flow of adsorption layers at oil/water and air/water interfaces. Interfacial rheology can be further divided into shear rheology and dilational rheology. Shear rheology involves shearing deformations of a surface element while keeping the area constant, whereas dilational rheology involves dilation/compression of the surface element which gives changes in area but a retained shape. (Erni et al. 2007)
17 The dilational rheology is a measure of interfacial resistance to changes in area. The interfacial dilational modulus, E is defined as the increase in surface tension for a small increase in area of a surface element. (Benjamins 2000)
E = dγ/d(lnA) (eq. 1)
γ = surface tension, A = area of the surface element
This is for the simplest case, when the interface is pure elastic (Wang et al. 2004). However, the interface will always exhibit some viscous behavior because of inter- or intramolecular rearrangements (Capek 2004). By compression/dilation of the interface at a given frequency, the elastic and viscous parts can be measured at the same time
(Benjamins 2000). Then E can be expressed as a complex function of frequency.
E = E’+ E” = Ed + iωηd (eq. 2)
where E’ is the storage modulus, E” is the loss modulus, Ed is the dilational elasticity, ω is
the angular frequency and ηd is the interfacial dilatonal viscosity. E’ is the real part and
represents the elasticity and E” is an imaginary part that accounts for the energy loss in the relaxation and can be expressed as the product of viscosity and the frequency. (Wang et al. 2004)
In this thesis, a dynamic drop tensiometer (se section 2.1.1.) has been used. The drop tensiometer can measure interfacial tension and complex modulus of an air/oil drop. Proteins form elastic films at the hydrophobic interface whereas OSA starch does not (Erni et al. 2007). By studying the change of these two parameters after the addition of OSA starch to a hydrophobic drop covered with livetin, a possible exchange of adsorbed livetin can thereby be observed.
18
2. MATERIALS AND METHODS
Instruments used in this thesis were a pH – electrode, an analytic balance, a dynamic drop tensiometer, a spectrophotometer, a homogenizer, a light diffraction apparatus for
measuring particle size and a zetasizer. The dynamic drop tensiometer, the light diffraction apparatus and the zetasizer are briefly described in this section.
2.1. Dynamic Drop Tensiometer
The dynamic drop tensiometer in this thesis is used to measure the interfacial tension and the complex modulus of an air drop during the adsorption of macromolecules. A schematic structure of the tensiometer is shown in Figure 4.
Figure 4. Schematic presentation of the drop tensiometer. a, function generator ;b, amplifier ;c, piezoelectric actuator; d, syringe; e, light source; f, cuvette containing the continuous phase; g, air/oil drop; CCD camera ;i, computer for data analysis; j, screen with a image of the drop. The cuvette and syringe is connected with a heating/cooling system for temperature regulation. (Erni. et al. 2007)
The point of the syringe is lowered into a hydrophilic liquid, corresponding the continuous phase. A drop of air or oil is then formed at the top of the syringe. During the adsorption of molecules/particles, the formed drop goes through sinusoidal oscillations of the volume. By registration of the drop image with a camera, the contour of the drop can be fitted to the Laplace equation and the interfacial tension and complex modulus can be achieved. (Erni et al. 2007)
19
2.2. Coulter light scattering apparatus
The light scattering apparatus used is based on laser diffraction measurements. Particle size distributions are achived by measuring the scattering intensity based on scattering models.
(Beckman Coulter 2006) A schematic structure of the apparatus is shown in Figure 5. The models are based either on the Fraunhofer theory (diffraction) or the Mie theory (scattering). In this thesis the Mie theory is applied. Unlike the Fraunhofer theory, Mie includes the refractive index of the particles studied. False results, caused by secondary scattering effects that may exist when small transparent particles are in the sample, can thereby be eliminated.(Teipel 2002)
Figure 5. A monochromatic beam from a light source passes through a number of optical components and the conditioned beam then reaches the sample. The particles in the sample scatter the light and create unique angular scattering patterns. These patterns go through Fourier transformation and are transformed into intensity patterns. The intensity patterns are detected by a photodetector array and the photocurrent is then processed into intensity flux pattern. Finally, a computer uses scattering theories and transforms the intensity flux patterns into particle size distributions. (Beckman Coulter 2006)
2.3. Zetasizer
When particles have a surface charge, for example caused by ionization or adsorption of charged molecules, the distribution of ions in the solution will be affected. An electric double layer, with a distribution that differs from the bulk solution, will surround the particle. It is the mainly this layer that affects particle interactions, not the charge at the interface. When the particle moves, the electric double layer comes along. The potential at the plane where the layer moves past the bulk solution is called the zeta potential.
(Malvern Instruments 2009)
Particles with high zeta potentials (conventionally <-30mV and >30mV) of the same charge repel one another and improve the stability. The zeta potential is affected by
20 surrounding elements, such as ion strength, and it is difficult to compare potentials
between different systems. However, by measuring the zeta potential it is possible to see how additives affect the charge stability of a specific emulsion. High potentials (both positive and negative) correspond to high stability. (Malvern Instruments 2009) With the Zetasizer it is possible to obtain the zeta potential through electrophoresis. In these measurements an electric field is applied across the dispersion which causes the particles to move towards the anode or the cathode. The velocity of the particles is then measured through laser Doppler anemometry. Interaction between particles and laser beam causes a scattering of the light in which oscillations are dependent of the velocities.
(Malvern Instruments 1991) From the velocities the mobility can be calculated (Malvern Instruments 1991) and from the mobility the zeta potential can be obtained from the Smoluckowski equation:
µ = ε ζ 4πη
where ε is the permittivity an η is the viscosity of the dispersion medium (Ducel et al. 2004).
21
2.4. Chemicals
The OSA-starch used was of waxy barley origin containing 7% amylase and 93%
amylopectin and was provided by Lyckeby Stärkelsen. The α-β-livetin fraction of the egg yolk had been purified from hen´s eggs. The eggs were manually broken and gently rolled onto a filter paper to remove remaining albumen according to the method of Le Denmat et al. (2000). This was followed by stepwise precipitation with NaCl and centrifugation, according to the method of McBee and Cotterill (1979), to receive the α-β-livetin fraction. The purification of the α-β-livetin fraction was made by a former employee at LTH. Other chemicals used throughout the thesis were of analytical grade.
2.5. Preparation of solutions
Solutions of 1 M HCl, 1 M NaOH and 1M NaCl were prepared. Buffers containing equal parts, 10 mmol/L each, of acetic acid, imidazole and Tris to a total concentration of 30 mmol/L were prepared in the pH range from 4.0 to 8.0. The ionic strength in the buffers was calculated with mass balances and equilibrium concentrations, and the ionic strength was then adjusted by the addition of different amounts of NaCl so that all the buffers achieved a total ionic strength of 0.038 M. An additional buffer at pH 9.0 and 6 mmol/L was prepared and used for the dissolution of the freeze-dried protein sample and the OSA-starch. All the buffers also contained 20 ppm sodium azide to prevent bacterial growth. 2.6. Titration curves
Solutions of OSA-starch and α-β-livetin with the concentration of 5g/L each were prepared with milli-Q-water. The protein sample was stirred for 30 min at room temperature and was then stored in the refrigerator overnight. The starch sample was placed in a boiling water bath under stirring for 10 min and was then left at room temperature over night. Solutions of NaOH and HCl with the concentration 0.01 M were prepared from
stocksolutions with the concentration of 1 M (se section 2.5.). Traditional titration curves were then made by acidification of the solutions (30 ml) through addition of 0.1 ml 1M HCl after which additions of 0.5 ml at the time was made. The pH-value was then plotted as a function of mmol/L NaOH. A two-way titration was also made by titration with acid to one solution and base to another. The curves were then put together.
22
2.7. Interactions in solution
2.7.1 Transmission profiles
Stocksolutions of α-β-livetin and OSA starch with a concentration of 50g/L and 20 g/L respectively were prepared in the same way as described in section 2.7.
The transmission of solutions with different OSA starch/ α-β-livetin ratios at a wavelength of 590 nm was then measured as a function of time with a Beckman DU®-70
spectrophotometer. pH values between 4 and 8 were investigated. The starch concentration was kept at 0.25% (w/v) whereas the α-β-livetin concentration was varied between 0% (w/v) and 1.25% (w/v). One sample containing only α-β-livetin at a concentration of 0.25% (w/v) was also analyzed.
2.7.2 Solubility
The solubility of the OSA starch and the livetin in solutions with different ratios between the macromolecules (as described in section 2.7.1.) was measured at pH 4.0, 4.5, 6.0 and 8.0. The solutions were centrifuged with a Minispin (Eppendorf) for 1 h at room
temperature. The amount of OSA starch in the supernatant was then determined by enzymatic degradation according to the method of Åman et al. (1994) and the α-β-livetin content was determined with the Bradford method (Bradford 1979). The supernatant was diluted 5 times for the starch determination and 12.5 times for the protein determination. 2.8. Interactions in emulsions
During this master thesis it was discovered that the egg yolk fraction was contaminated with a small amount of glucose. This does most likely not affect the earlier experiments, as the concentration was very low (≈ 30 ppm) but to be able to achieve adsorption isotherms the contamination had to be removed. The α-β-livetin was therefore dialysed and freeze dried.
2.8.1. Adsorption isotherms
Oil-in-water emulsions at pH; 4.0, 4.5, 6.0, and 8.0 were prepared with 2.5% (v/v) of medium chain triglyceride oil (MCT oil) Miglyol 812 F and a protein concentration of 0.1% (w/v). The emulsions were pre-mixed with a high shear mixer Ystral X10/25 for 3 min followed by high pressure homogenization in a lab-scale valve homogenizer at 13 MPa. The lab-scale valve homogenizer has been described elsewhere (Tornberg and Lundh
23 1978). The area-weighted average droplet size, d32 was determined by light diffraction
using a Coulter LS130. Each emulsion was divided into four tubes, to which different amounts of OSA starch were added. The emulsions were then equilibrated for 5 h. All experiments were performed at room temperature. After equilibration the emulsions were separated by mild centrifugation (Beckman GPR Centrifuge IMP 5899 and Eppendorf Minispin) in two steps to avoid coalescence. The OSA starch content in the subnatant was then determined by enzymatic degradation according to Åman et al. (1994) and the α-β-livetin content was determined with the Bradford method (Bradford 1976).
Reference samples of protein and starch containing no dispersed phase went through the same step of homogenization, centrifugation and equilibration as the emulsions. The adsorbed amount of OSA-starch was then obtained from the difference between the amount in the reference samples and the amount in the subnatant of the separated emulsion:
cadsorbed= creference- csubnatant (eq. 5)
The surface load, Г, was calculated from the adsorbed amount and the specific surface area of the emulsion:
Г = cadsorbed d32 (eq. 6)
6Φ
where d32 is the area-weighted droplet diameter and Ф is the dispersed phase volume
fraction.
To see if the adsorbed amount of protein was altered by the starch addition, the difference in csubnatant of α-β-livetin between an emulsion only containing livetin and an emulsion
with both livetin and OSA starch (mass ratio of 1/1) was studied. The emulsion preparation and the protein determination were made in the same way as previously described in this section.
24
2.8.2. Analyzing α-β-livetin and OSA-starch solutions with the drop tensiometer
Stocksolutions of α-β-livetin and OSA starch with a concentration of 25g/L were prepared in the same way as described in section 2.5. Solutions containing 1g/L α-β-livetin were then made in the pH range from 4 to 8. The samples were analyzed with the drop
tensiometer at a period of 10 s, a droplet area of 15 mm2 and deformation amplitude of 2. Amplitudes between 1 and 5 were investigated; 2 was chosen as it gave the maximal signal. The protein solutions were analyzed in the drop tensiometer at gentle stirring for 2200 s*, after which an addition of starch was made, so that the weight proportion between
the macromolecules became 1/1. The solution was further analyzed until 7200 s (2h) had passed. The protein solutions were also analyzed without addition of OSA-starch for 2 h.
2.8.3. Emulsion stability
Oil-in-water emulsions at pH; 4.0, 4.5, 6.0, and 8.0 were prepared in the same way as in section 2.8.1. The emulsion was divided in two, and OSA-starch was added to one half of the emulsion. The emulsions were then transferred into glass tubes. The tubes were
overfilled to prevent coalescence and put in a turning wheel where they were turned upside down at a slow pace to avoid creaming. The amount of flocculation and coalescence in the emulsions were observed by particle size distribution and microscopic analysis after 1 h, 1 day, 3 days and 7 days.
Microscopic images of the 10-fold diluted emulsions were taken with an Olympus light microscope BX60 with an Olympus U-CMAD-2 camera at magnifications of 200× and 500×. Five photos at each magnification were taken. The area-weighted droplet size, d32
was determined by light diffraction using a Coulter LS130. The Mie-theory was applied whit the real part of the refractive index 1.47 and the imaginary part 0.00. The imaginary refractive index expresses the adsorption of light, and since organic molecules do not adsorb the laser light, the refractive index was set to zero (Malvern Instruments 1991). The zeta potential of the emulsions were also measured after 1 h with a Zetasizer 4 (Malvern Instruments). The particle size measurements and the measurements with the zetasizer were made in triplicate.
25
3. RESULTS
3.1. Titration curves
Titration curves of α-β-livetin and OSA starch were made by addition of HCl/NaOH as described in the methods section. The titration curves of α-β- livetin and OSA starch are shown in figure 7A and 7B respectively. The two-way titration worked, and the curves fit well together. Therefore these are the ones that are discussed because no consideration is needed to be taken to the acidification. The curve for α-β-livetin indicates that the average pI (the pH where the net charge is zero) for the solution is around six. The OSA starch can only attend a negative charge. From the titration curve it seems to become negatively charge at a pH around 4, as the curve begins to incline at this value (= titration of groups).
A B
Figure 7. Two-way titration curves, A; α-β- livetin, B; OSA starch
To get a better a better picture of the charges of the macromolecules the amount of titrated groups were calculated by subtraction of the pH after the addition of acid or base from the added amount. The charge/glycopyranosyl unit of starch and charge/amino acid where then held as a function of pH, se Figure 8 and 9. For this calculation, the average pI of livetin was assumed to be 6.1, as this is in line with the calculated average pI from the theoretical and observed pI values reported by Nilsson et al. (2006) and also with the average pI indicated by the titration curve. No consideration was taken for the different amounts of the proteins in the fraction since the relative concentrations have not been quantitatively determined. The results indicate that the protein solution has a higher positive charge at low pH values, as opposed to negative charge at high pH values. The interval where the net charge is zero seems to be quite large (≈ 6 – 7.5), and the negative charge density is low
26 even at pH 8. For the starch, the difference in charge/glycopyranosyl unit seems to be quite small between high and low pH values.
Figure 8. Charge/amino acid as a function of pH. Figure 9. Charge/glycopyranosyl unit as a function of pH.
To be able to discuss the charge/amino acid the total amount of the different amino acids in the α-β-livetin fraction was calculated from the protein sequences achieved from the Expasy protein database. As in the calculations for the charge/ amino acid curve, no consideration was taken for the different amounts of the proteins in the fraction. The amino acids with acidic side chains; aspartic acid and glutamic acid, consists of 12.7 % of the total amino acid content, whereas the ones with basic side chains; arginine, cystein and lysine consists of approximately 15.4 %. Histidine, with its neutral side chain, consists of 2.1 %.
Table 2. The amino acid content of α-β-livetin.
*The pKa values for side chains in proteins, taken from Damodaran (1996) Amino Acid pKaR* % of total
amino acid Alanine 7,2 Arginine >12 3,7 Aspargine 3,8 Aspartic acid 4,6 6,2 Cysteine 8,8 4,8 Glutamine 4,4 Glutamic acid 4,6 6,5 Glycine 6,8 Histidine 7,0 2,1 Isoleucine 4,2 Leucine 7,4 Lysine 10,2 6,9 Methionine 1,5 Phenylalanine 2,6 Proline 4,4 Serine 8,2 Threonine 6,1 Thryptophan 1,1 Tyrosine 2,7 Valine 7,4
27
3.2. Interactions in solution
3.2.1 Transmission profiles
At the preparation, the OSA starch solution was a bit hazy, whereas the livetin solution was almost clear. Therefore, the livetin concentration was the one to be varied whereas the OSA starch concentration was held constant. Figure 10 shows the transmission profiles of different ratios of livetin/OSA starch at pH 4.
Figure 10. The relative transmission, compared to the buffer, as a function of the livetin to OSA starch mass ratio. All mixtures were prepared with a buffer solution at pH 4.0 and contained 0.25 % (w/v) OSA starch and a concentration of livetin between 0 % (w/v) and 1.25 % (w/v). Open symbols represents the relative transmission at time 0 and closed symbols represent the transmission after 155 minutes.
The low transmission value at a livetin/OSA starch ratio of 2/1 indicates the presence of a large amount of livetin-OSA starch complexes. The complexes seem to be formed quite fast, since the transmission value at this ratio is lower even at time zero. A significant decrease in transmission was observed at a ratio of 1/1 as well.
In table 3 and figure 11, one can observe an increase in the transmission after 1000 minutes of the solution containing livetin and OSA starch at a ratio of 2/1. This indicates
sedimentation of the complexes. An increase in transmission was also seen for the solution only containing livetin. The transmission of the other solutions still decreased slightly after 1000 minutes.
28 Table 3. Transmission values of solutions with different livetin/OSA starch ratios in a time scale of zero and 1000 minutes.
Figure 11. Transmission values as a function of time for different livetin/OSA starch ratios at pH 4.
For the other pH values, see figure 12, no strong indication of complex formation was observed. Although at pH 4.5, a small decrease in transmission was seen at the ratios of 2/1 and 1/1. Time (min) 1/0 SD 0/1 SD 5/1 SD 2/1 SD 1/1 SD 1/2 SD Blank SD 0 93,75 ±2,49 91,21 ±2,06 90,62 ±0,88 83,18 ±0,13 88,34 ±1,39 90,16 ±2,53 99,29 ±0,74 5 92,48 ±2,67 90,59 ±2,10 90,51 ±0,76 79,94 ±0,17 87,02 ±0,77 89,65 ±1,54 99,34 ±0,58 15 91,39 ±2,65 90,28 ±1,92 90,48 ±0,58 75,22 ±0,75 85,28 ±0,74 89,10 ±1,50 99,74 ±0,42 25 91,23 ±3,72 90,64 ±3,01 90,63 ±0,01 70,76 ±0,49 84,15 ±0,49 88,84 ±1,26 99,50 ±0,53 35 91,15 ±4,31 90,49 ±2,74 90,95 ±0,71 66,82 ±0,19 82,87 ±1,15 88,15 ±1,73 99,64 ±0,40 95 89,31 ±2,60 90,21 ±2,49 89,66 ±0,71 52,69 ±0,16 79,71 ±1,30 87,60 ±1,85 100,19 ±0,35 155 88,83 ±2,30 90,15 ±2,53 89,10 ±0,71 46,09 ±0,35 78,09 ±1,19 87,29 ±1,92 100,21 ±0,34 1000 95,88 ±1,39 89,59 ±2,43 85,76 ±1,82 66,54 ±0,62 74,57 ±0,93 86,45 ±1,53 100,37 ±0,03
29
8.0
A B C
Figure. 12 The relative transmission, compared to the buffer as a function of the livetin to OSA starch mass ratio. The mixtures were prepared with buffer solutions at pH 4.5 (A), 6.0 (B) and 8.0 (C) and contained 0.25 % (w/w) OSA starch and a livetin concentration between 0 and 1.25 % (w/v).
The low transmission at pH 4.0 for livetin and OSA starch, observed in Table 4 and 5, indicates that the solubility of the macromolecules is limited at this pH value. The solubility of livetin is improved with increased pH values, whereas the solubility of the OSA starch seems to be almost the same between pH 4.5 and 8.0, which is also reflected by the charge density in Figures 8 and 9.
Table 4. The transmission of a solution containing 0.25 % (w/v) livetin after 155 minutes.
Livetin pH value Transmisson (%) SD 4,0 88,83 ±2,30 4,5 93,71 ±1,26 6,0 98,17 ±3,02 8,0 100,00 ±0,08
Table 5. The transmission of a solution containing 0.25 % (w/v) OSA starch after 155 minutes.
OSA starch pH value Transmisson (%) SD 4,0 90,15 ±2,53 4,5 95,34 ±0,11 6,0 94,05 ±2,28 8,0 95,05 ±0,88
30
3.2.2 Solubility
In order to investigate how the addition of starch affected the solubility of livetin, solutions with different livetin/OSA starch ratios were centrifuged and the concentrations of the macromolecules in the subnatant were analyzed. The large standard deviations for the protein concentrations made it difficult to draw any conclusions of the results. Although, one interesting observation can be made; at pH 4.0, the solubility of the OSA starch is much lower at the ratio 1/1 of livetin and OSA starch than for all the other solutions (Table 6).
For the OSA starch determination in the subnatant, the recovery was quite low, only around 70%. Though, the method of Åman et al. (1994) was developed for unmodified starches, and the lower recovery for the OSA starch was probably due to incomplete hydrolysis (Nilsson and Bergenståhl 20062). By assuming that the recovery is equal in the different samples, comparison between the starch containing samples can still be done and the error becomes minimal.
Table 6. Concentrations of livetin and OSA starch in the subnatant of centrifugated solutions containing different ratios of the two macromolecules.
pH value
Livetin/OSA starch ratio (w/v)
Livetin only (g/l)
Livetin (with OSA starch)
(g/l)
OSA starch (with livetin) (g/l) 4.0 0/1 − − 1,81 ± 0,12 5/1 9,49 ± 0,09 10,89 ± 1,08 1,79 ± 0,01 2/1 4,21 ± 0,64 4,15 ± 0,60 1,83 ± 0,14 1/1 2,60 ± 0,41 2,27 ± 0,57 1,07 ± 0,28 1/2 1,57 ± 0,55 1,51 ± 0,78 1,88 ± 0,20 8.0 0/1 − − 1,73 ± 0,01 5/1 10,39 ± 0,65 10,72 ± 1,29 1,79 ± 0,00 2/1 5,51 ± 0,74 4,83 ± 0,45 1,85 ± 0,01 1/1 2,82 ± 0,85 2,72 ± 0,87 1,82 ± 0,06 1/2 1,98 ± 0,28 1,67 ± 0,64 1,82 ± 0,07 3.3. Interactions in emulsions 3.3.1. Adsorption isotherms
Oil-in-water emulsions were produced with livetin, and different amounts of OSA starch were added after the emulsification. The area-weighted droplet diameter, d32, varied
31 sensitive to flocculation. The large particle sizes observed for the low pH values could therefore be due to flocculation. However, the particle size measurements were made immediately after the emulsion formation and the curves achieved for the particle diameters had a Gaussian distribution.
An interesting observation was made after the addition of starch to the emulsions at pH 4.0. A greater degree of creaming could be visually observed in the starch containing emulsions, especially at the highest OSA starch concentration (1g/l). The same observation was made at pH 4.5, though it was not as distinct as at pH 4.0.
Table 7. Results from the adsorption experiments.
OSA starch Livetin
co (g/L) App. ceq (g/L) Γ (mg/m2) Adsorption yield (%)** (g/L) co App. ceq (g/L) Γ (mg/m2) Adsorption yield (%)** pH value d32 (µm) 4.0 4,78 0,09 0,06 0,46 0,16 1,00 0,33 11,47 0,36 0,28 0,19 0,31 0,03 1,00 0,25 14,12 0,44 0,46 0,31 0,16 0,01 1,00 0,26 13,56 0,42 0,93 0,62 -0,12* 0,00 1,00 0,25 13,89 0,43 4.5 8,77 0,09 0,06 0,13 0,00 1,00 0,15 30,13 0,51 0,28 0,19 1,28 0,02 1,00 0,14 30,63 0,52 0,46 0,32 1,38 0,02 1,00 0,16 29,76 0,51 0,93 0,64 1,73 0,03 1,00 0,18 28,75 0,49 6.0 2,52 0,09 0,06 0,17 0,11 1,00 0,72 2,28 0,13 0,28 0,19 0,40 0,08 1,00 0,71 2,44 0,14 0,46 0,32 0,81 0,10 1,00 0,69 2,79 0,16 0,93 0,63 0,57 0,04 1,00 0,73 2,26 0,13 8.0 2,00 0,09 0,065 0,05 0,04 1,00 0,79 1,05 0,08 0,28 0,195 0,11 0,03 1,00 0,80 0,87 0,06 0,46 0,325 0,29 0,05 1,00 0,80 0,90 0,07 0,93 0,65 0,20 0,02 1,00 0,82 0,62 0,05
*At concentrations as high as 0.93 g/L, errors by the subtraction of large values may occur. This surface load may thereby be assumed to be zero. Errors may occur for the surface loads for all the higher concentrations. ** The adsorption yield is defined as the quota between cads and c0.
Very high surface loads of livetin were obtained at pH 4.0 and 4.5 whereas the surface loads at pH 6.0 and 8.0 were what could be expected for a monolayer of a water soluble protein. (Walstra1996) As can be seen in figure 13A, a minimum in created emulsion surface area by the emulsification is seen at pH 4.5. Conversely, this pH corresponds to the
32 highest surface loads (figure 13B). For all pH values, the surface area is negatively
correlated with the surface load of livetin.
A B
Figure 13. Emulsions with a ratio 1/1 of livetin/OSA at different pH values. The plots shown are; a, the emulsion surface area versus pH, b; the surface load of livetin versus pH and c, the surface load of OSA starch versus pH.
The surface loads obtained for OSA starch (figure 14) ranged between zero and 1.73 mg/m2. The highest surface load was obtained at pH 4.5 followed by pH 6.0. Low
adsorption was observed at pH 4.0 and 8.0. At pH 4.0 the adsorption of OSA starch
decreased linearly (R2=0.9946) with increasing apparent equilibrium concentration.
Figure 14. Adsorption isotherms for the adsorption of OSA starch at livetin at different pH values.
In order to see if the adsorbed amount of protein was altered by the starch addition, the difference in apparent equilibrium concentration (ceq) of the livetin in an emulsion only
33 containing livetin, as well as an emulsion with both livetin and OSA starch were
investigated at pH 4.0 and 8.0. No substantial differences between the apparent equilibrium concentrations were observed.
3.4. Interfacial rheology
3.4.1. Analyzing α-β-livetin and OSA-starch solutions with the drop tenisometer Figure 15A shows the interfacial tension of α-β-livetin solutions at different pH values. The highest interfacial tension can be observed at pH 8.0 and it then decreases with pH value. This indicates a greater adsorption at low pH values. The complex modulus for the solutions is shown in figure 15B.
A B
Figure 15. Interfacial rheology of a surface covered with livetin. A, interfacial tension as a function of pH; B, complex modulus as a function of pH.
The effect of starch addition to the livetin solution was also investigated, se figure 16. The result indicates that the OSA starch lowers the complex modulus of the interfacial layer.
A B
Figure 16. The change in interfacial rheology for a surface covered with livetin caused by the addition of starch. The addition was made after 2200 seconds. A, interfacial tension as a function of pH; B, complex modulus as a function of pH.
34
3.5. Emulsion stability
To see how the stability was affected by the starch addition, a stability study was made. An indication of an increased particle diameter for the starch containing emulsions can be seen at pH 4.0 (figure 17A), though the differences were not significant. At pH 4.5 (figure 17B), the standard deviations were large and it is thus hard to make any conclusions. As for the higher pH values, 6.0 and 8.0, see figure 17C and 17D, no trend for the particle sizes could be observed.
A B
C D
Figure 17. The area-weighted average droplet size of the emulsions after 1h, 1 day, 3 days and 7 days. The open bars represent emulsions only containing livetin whereas the closed bars represent emulsions with both
livetin and OSA starch. A, pH 4.0; B, pH 4.5; C, pH 6.0 and B, pH 8.0.
The emulsions can be considered quite stabile towards coalescence in the time scale investigated. The zeta potential of the emulsions was measured 1h after the emulsification (Figure 18). The measurements show that the pI of the emulsion droplets lies around pH 4.5. At lower pH values the droplets are positively charged and at higher pH values they
35 are negatively charged. The addition of starch lowered the zeta potential of the emulsions and the decrease was significant at pH 4.0 and 4.5.
Figure 18. The zeta potential versus the pH of the emulsions. Open symbols represents the emulsion only containing livetin whereas closed symbols represents emulsions with both livetin and OSA starch.
No significant differences were observed between the microscopic images of emulsions only containing livetin and emulsions with both livetin and OSA starch. High flocculation and some coalescence was, however, observed for all emulsions at pH 4.0 and 4.5 (Figure 19), whereas no flocculation or coalescence could be observed at pH 6.0 and 8.0 (Figure 20).
Figure 19. Microscopic image of an emulsion with Figure 20. Microscopic image of an emulsion at livetin at pH 4.5, magnification ×500.7 days after pH 6.0 at magnification ×500, 7days after emulsion
36
4. DISCUSSION
4.1. The charge of the macromolecules
The two-way titration curves indicated that the pI is around six for livetin. This is in line with the average pI calculated from theoretical and observed pI values for the proteins observed by Nilsson et al. (2006). As seen by the charge/amino acid, an indication of a higher charge density at low pH values than high, can be seen. Arginine, histidine and lysine have a positive charge at pH values below 4.6. About 13% of the amino acids would therefore be likely to be positively charged at low pH values, however, as seen by the charge/amino acid, around 20% of the amino acids are positively charged (pH≈ 3). Since no considerations have been taken for the different amounts of the proteins in the fraction, this may partly explain the deviation. At pH 6, about 13% of the side chains are positively charged, and 13% are negative. The net charge will therefore be zero. This is in line with the charge/amino acid curve. Cysteine becomes negatively charged above pH 8.8. At this pH aspartic acid, glutamic acid and cysteine will be negatively charged, whereas arginine and lysine still are positively charged. About 18% will thereby be negatively charged whereas 11% will have a positive charge. The amino acid content explains the lower charge density at high pH values that can be observed in the curve.
From the titration curves, it seems that the OSA starch gets a negative charge around 4.0. The charge is almost the same in the entire pH interval, according to the
charge/glycopyranosyl unit curve, only slightly lower at pH 4.0. This is in line with the transmission profile achieved for the starch. The transmission value was lower at pH 4.0 but then almost the same between pH 4.5 and 8.0.
The charge/glucopyranosyl unit does not reach the degree of substitution, 0.0213 even at pH values as high as 8.0. No explanation for this can unfortunately be given.
4.1. Interactions in solution
The results of the transmission profiles indicate that a large amount of livetin/OSA starch complexes are formed at pH 4.0, especially at a ratio of 2/1 but also at a ratio of 1/1. This could be due to electrostatic or hydrophobic interactions or perhaps a combination of both. As shown by Nilsson et al. (2006) and also indicated by the results from the titration, all the protein species have a pI higher than 4.0 and the proteins are thereby positively charged at this pH value. The transmission profile for livetin at pH 4.0 indicates low solubility, and
37 with time, sedimentation of the protein. This can be due to the formation of protein
aggregates or partial denaturation of the proteins. For a solution that only contains one protein species, aggregation occurs around the isoelectric point. At pH values below or above the pI, the proteins repel each other electrostatically, whereas they can come close and aggregate at the pI. (McClements 1999) If the lower transmission was due to the formation of aggregates, at least one of the protein species would be expected to have an isoelectric point near 4.0. However, this was not indicated by the results of Nilsson et al., nor by the results achieved in this study. Thereby protein denaturation is a more likely explanation. This occurs at extreme pH values because sensitive internal parts of the protein have the same charge and they repel each other, or as charges used for stabilization are removed. (Scopes 1994) The protein thereby becomes partly or fully unfolded and forced to have some hydrophobic parts towards the solution. The solubility is decreased and the protein may precipitate. The pKa value for the OSA starch can be expected to be somewhere around 4.76, as this is the pKa value for acetic acid. At pH 4.0, the starch may thereby have some negative charges, though the charge density is likely to be relatively low. This is further indicated by the titration curves for the OSA starch and the lower solubility of the OSA starch at pH 4.0 compared to the other pH values investigated. The low charge density of OSA starch at pH 4.0 and the fact that some species in the livetin may be partly denaturated makes hydrophobic interaction likely. The poor solubility of livetin and OSA starch makes it energetically favorable for them to interact and turn their hydrophobic parts away from the hydrophilic solution. However, since the macromolecules have different charges, electrostatic interactions may also occur and it is therefore hard to distinguish one effect from the other.
The transmission profiles for solutions with different livetin/OSA starch ratios at pH 4.0 indicates the formation of livetin-OSA starch complexes, especially at a ratio of 2/1 but also at a ratio of 1/1. A small indication of complex formation was also seen at the ratio of 1/2. The fact that the livetin fraction consists of different species in different quantities makes it difficult to say why the largest amount of complexes seems to be formed at a ratio of 2/1. Quite interesting, though, Neirynck et al. (2007) observed the same maxima for the interaction between whey protein and pectin. They also noted the same patterns for the other ratios between the macromolecules, although the transmission values observed were lower for the whey protein/pectin solutions. The complex formation, or at least the amount
38 of insoluble complexes formed, thereby seemed to be higher in the case of whey
protein/pectin interactions. In the experiments done in this thesis, no interaction was observed at a ratio of 5/1 between livetin and OSA starch in solution. However, the macromolecule concentration is very high (1.25 % (w/v)) at this ratio, which gives a high viscosity, less movements of the particles and therefore less interaction. Thus, this point must be seen as an extreme. Unfortunately, such high concentrations were needed to be used because of the turbidity of the pure OSA starch solution. Neirynck et al. were able to use lower concentrations of the macromolecules and saw an interaction at the ratio of 5/1, however it was weaker than the interactions at all other ratios. These authors did not propose any theory for the differences in the amount of formed complexes at the different ratios; unfortunately, no explanation can be given in this thesis either.
When studying the transmission of the complexes/macromolecules over time, an indication of sedimentation was observed in a solution containing pure livetin, but no sedimentation was observed for the solutions containing livetin and OSA starch at the mass ratios of 5/1, 1/1 or 1/2. This indicates that the addition of OSA starch decreases the sedimentation of the protein at these ratios, at least in the time span investigated. However, the difference in total macromolecule concentration and the transmission value of the solutions at the beginning of the experiment may affect the results. Sedimentation was observed also at a livetin/OSA starch ratio of 2/1, and the transmission value of this solution was higher than for all the other solutions.
The solubility experiment did not indicate an increase of the solubility of the protein induced by the OSA starch addition, nor did it show any indication of the opposite. Because of the large standard deviations for the concentrations of protein in the subnatant, it is difficult to make any conclusions at all. The large standard deviations are most likely due to errors caused by too small volumes used for the dilution of the solutions. From the large complex formation at the ratio 2/1 between livetin and OSA starch at pH 4.0, and observed decrease in the transmission with time; one would expect to see the largest sedimentation of OSA starch and livetin at this ratio in the solubility measurements. No decrease could though be observed at a ratio of 2/1; however, an increased sedimentation of the starch was seen at a ratio of 1/1 at pH 4.0. It is unfortunate that no conclusions could be drawn from the results for the protein concentrations, since this makes it more difficult to interpret the increased sedimentation of the starch. The decrease in the OSA starch
