BUILDING VISCOELASTICITY IN FOODS THROUGH INTERFACES AT DIFFERENT LENGTHSCALES

BUILDING VISCOELASTICITY IN FOODS THROUGH INTERFACES AT DIFFERENT LENGTHSCALES

Raffaele Mezzenga

Department of Physics, University of Fribourg, Perolles, Fribourg, CH1700 Switzerland (raffaele.mezzenga@unifr.ch)

Nestlé Research Center, Vers-Chez-Les-Blanc, Lausanne 26, CH-1000, Switzerland (raffaele.mezzenga@rdls.nestle.com)

In the present talk we discuss how the viscoleastic properties of foods can be tuned by suitably designing interfaces at different length scales. Two extreme cases will be discussed; in the first example the interface providing viscoelastic behaviour to the food systems are the water-lipid interface in self-assembled liquid crystalline phases. Depending on the specific organization of the liquid crystalline structure, the interface, always structured in the typical length scale of 5-10nm, may exhibit plastic (lamellar phase), viscoelastic (hexagonal phase) or strongly elastic behaviour (cubic phases). As a consequence, the rheo-perceptive properties of the corresponding food can be highly varied.

In the second case we discuss high internal phase emulsions formed by oil-in water emulsions stabilized by crosslinked polymer films. In this case, the rheological properties can be directly tuned by both the average radius of droplets, as well as interfacial tension of the protein film or crosslink density.

1. INTRODUCTION

The most effective way to tune rheological properties of foods is to efficiently design and control the viscoelatic interfaces at different characteristic length scales. On the very low length scale limit, lyotropic liquid crystals offer a very valuable route.1-5 Systems formed by lipids and water are a representative example of this category of materials, and are of interest for a multitude of different applications such as dispersion technology, cosmetics, pharmaceutical products, encapsulation systems, foods, etc.6-13 

Since different types of lipids result in slightly different phase diagrams, extensive efforts have been made in order to explore all possible structures. In systems formed by monoglyceride and water, on which we will focus in the present paper, the extent of each specific phase region depends mostly on the length and degree of unsaturation of the hydrophobic chain.6,14-16 The most common types of liquid crystalline structures observed in monoglyceride-water systems are lamellar phases (Lα and Lβ), the hexagonal columnar phase, and two types of bicontinuous cubic phases, the gyroid (Ia3d) and double diamond (Pn3m). Together with the latter two structures a third cubic phase, the primitive (Im3m) is occasionally observed.17,18 

Whereas several decades of investigation on the morphology of lyotropic liquid crystals have established reliable methods for identification and assessment of their structures, there have been few rheological studies on these complex fluids.  To date, a specific rheological signature has not been established for all of the known self-assembled nanostructures in surfactant-water binary systems, and the viscoelastic and plastic mechanisms regulating their response to stress and deformation is still being debated.

On the opposite extreme of length scale, emulsion and dispersion technology have been both widely employed to design functional viscoelastic foods. Eventually, high internal phase oil-in-water emulsions (HIPEs) have been designed, where the balance of elasticity versus viscous properties can be tuned at some extent by controlling the volume fraction of the continuous phase, that is the amount of total interfaces. In general, however, the increase in elastic modulus for these blends generally remains rather limited even when the volume fraction of the continuous phase is reduced below 10% (Mason et al., 1995).19 Further reduction of the continuous phase volume fraction, generally results in coalescence of the emulsions under shear (Dimitrova and Leal-Calderon, 2004).20  A sharp increase in elasticity of the oil can also be reached by hydrogenation of the liquid oil. However, in addition to irreversibly modifying the oil phase, this also corresponds to a strong increase of viscosity of the oil phase, with a subsequent reduction of processability.

In the present work we describe how a careful design of interfacial properties from nanometer to micrometer scale can result in affecting the overall rheological properties of foods. A selected set of examples will be discussed.

2. EXPERIMENTAL SECTION

2.1 Materials

Commercial monolinolein Dimodan U/J, was a generous gift from Danisco (Brabrand, Denmark) and was used as received. Deionized water was used for preparing lipid-based liquid crystalline structures. The sample preparation is described elsewhere.21  For emulsions and HIPE studies, β-Lactoglobulin (β-Lg, Mw=18·103 g/mole), was used as emulsifying protein (supplied by Davisco Foods Inc.). The cross-linking behavior of β-Lg in solution, once this was absorbed on the oil droplets interface was achieved by either thermal or chemical pathway. 22 

2.2 Methods

Rheological Measurements

The rheological signatures of the various liquid crystalline phases and emulsions were determined using a Paar-Physica MCR 500 rheometer in strain-controlled mode (Direct Strain Oscillation). The measurement cell was a DIN concentric cylinder or parallel plate for HIPE. For the liquid crystalline phases, in order to minimize water loss by evaporation, a solvent trap system using concentric Teflon disks and a low viscosity sealing oil were used. The contribution of the solvent trap to the overall torque measurement is below the threshold detection limit of the instrument (<0.1 μNm). Water losses were measured and found to be negligible.

All measurements were performed in the linear viscoelastic regime.

In order to determine the phase transition temperatures, temperature sweeps were performed for 12%, 20% and 25% aqueous solutions with a heating rate of 0.2°C/min. The strain amplitude was 0.01% and the frequency 1 rad/s.

In order to determine the relaxation times, frequency sweeps were also performed.

3. RESULTS AND DISCUSSION

3.1 Liquid crystalline systems

Measurement of the storage and loss moduli as a function of the temperature reveals the major transitions that occur in the phase diagram between different liquid crystalline phases. Figure 1 shows an example of temperature scan for G’ and G’’ for the system with 20 wt% water content. The different LC phases and their boundaries are identified by separate SAXS experiments. The changes in G’ with temperature capture all of the transitions.  Between 20 and 50°C, that is within the double gyroid Ia3d phase, G’ decreases gently; however, at 50°C, in correspondence with the double gyroid-double diamond cubic phases transition (Ia3d-Pn3m transition), G’ starts increasing until the temperature of 62°C is reached. The increase of G’ in this temperature window is directly related to the Pn3m structure. A similar increase in G’ with temperature has been observed by Jones and McLeish in other systems,20 but the lack of SAXS data did not allow the authors to identify the different rheological behavior of the Ia3d and Pn3m. Between 62 and 70 °C, G’ decreases linearly, in line with the fact that in this temperature window, coexistence of the Pn3m and Hex phases is observed by SAXS. Beyond 70 °C, in the hexagonal phase region, a slight increase of G’ with temperature is observed up to 85°C, probably related to the decrease of the lattice parameter with temperature. Finally, at temperatures above 85 °C, G’ decreases sharply, eventually reaching a linear regime which suggests the coexistence of Hex and isotropic phases. This coexistence, is not easily detectable by other techniques, since cross-polarized microscopy, cannot distinguish between two phases among which one is birefringent, while in SAXS the HEX form factor partially covers that of isotropic fluid. Thus, a temperature scan of G’ allows such a transition to be more easily identified. Finally we observe that, apart from the bicontinuous cubic phase region, where G’’ grows monotonically with temperature, the behavior of G’’ in Figure 1 resembles very closely that of G’, and can be used to quantify the Pn3m-to-Hex and Hex-to-FI transitions but not the Ia3d-to-Pn3m one.[pic 1]

...