BUILDING VISCOELASTICITY IN FOODS THROUGH INTERFACES AT DIFFERENT LENGTHSCALES

...

Figure 1. Storage (G’) and loss (G’’) moduli as a function of temperature for the liquid crystalline systems formed by 80 wt% lipid and 20 wt% water. The temperature windows at which different phases are observed are obtained by separate SAXS analysis.

In a frequency scam experiments, one can probe the various liquid crystalline phase at different typical response times. Typically, the inverse of the frequency, ω, at which the crossover between G’ and G’’ occurs, τMAX ~ 1/ω, is called the longest relaxation time and that is the characteristic time at which the structured fluid relax back to the equilibrium configuration when perturbed by, for example, shear oscillations. At frequencies close to 1/τMAX the fluid is no longer viscous but rather exhibits a viscoelastic behavior. Above 1/τMAX the elastic modulus dominates and a rubbery plateau in G’ is obtained while G’’ may decrease to eventually reach a minimum before rising again. By further increasing the frequency, a leathery or higher transition crossover is observed. In this region the increase in G’’ is sharper than G’, owing to dissipation mechanisms or short relaxation time responses. Finally at very high frequencies, a glassy region is encountered where G’’ grows faster than G’.

Figures 2a-c depict the frequency scan experiments for the lamellar, hexagonal and double gyroid cubic phases, respectively. The three phases show very different viscoelastic behavior, and are characterized by very different relaxation times f the order of 100s. However, since at low frequencies, G’ is larger than G’’, this viscoelastic behavior corresponds to the so called leathery transition, indicating a plastic behavior associated with dissipation mechanisms. By further decreasing frequency to values lower than the minimal observable value in the present experiments (ω

Finally, Fig 2c. shows the viscoelastic signature of a double gyroid cubic phase. Because of the maximum in G’’ and the monotonic increase of G’, this viscoelastic signature is reminiscent of simple Maxwell behavior. Thus, in Figure 2c, the experimental storage and loss moduli are compared with multiple Maxwell elements expressed as:[pic 2]

[pic 3]

[pic 4]

[pic 5] (6b)

Figure 2. Storage (G’) and loss (G’’) moduli versus frequency for the lamellar (A), Hexagonal (B) and double gyroid Ia3d (C) phases.

It can be noted that, the real experimental behavior is much more complex than multiple Maxwell modeling. Also, the characteristic relaxation time τMAX > 10 seconds, indicate that these cubic phases are highly elastic and characterized by very slow and complex relaxation, associated with the degrees of freedom of the lipid-water interface.

3.1 HIPE systems

Figure 3 illustrates the morphology of a monodisperse oil-in-water emulsion template whose interfaces are stabilized by an absorbed thin film of protein (beta-lactoglobuline). Upon thermal crosslinking of the protein layer, the interfacial energy is increased, and elasticity is provided to the interfaces. Thus, when the water solvent is evaporated, the protein layer posses sufficient elasticity to survive the extensional dilatation of about 10% leading a closely packed emulsion template to a high internal phase emulsion (HIPE) in which the residual water content can be kept as low as 0.1 to 1%.

[pic 6]

Figure 3. From left to right: (i) Emulsion template (droplet radius 24 μm) (ii) Twofold-layered thin film of polyhedron gel obtained upon water evaporation on a glass substrate and (iii) 3D gel resulting from a monodispersed emulsion template, as revealed by confocal microscopy (the protein phase is labeled by fluorescent rhodamine probe).

Clearly, in these conditions, the overall viscoelastic properties of the resulting gel are mostly determined by two factors: (a) the size of the droplet template (which fixes the amount of interfaces present in the gel) and (b) the intrinsic elasticity of the protein film network, the latter mostly depending on the physical or chemical pathway followed to crosslink the interfaces.

Figure 4a illustrates the dependence between the shear modulus and the size of droplets. A G’~1/r scaling law is found, which states that the shear modulus can be efficiently increased by reducing the size of the droplet template. Thus, a droplet template of 1 μm correspond to a shear modulus of 20kPa, while 0.1 μm droplets can be used to produce viscoelastic gels with G’=0.1MPa.

The remarkable feature encountered with these materials is that, despite of the high elastic behavior, the original emulsion template can be recovered by re-hydrating the gels, as the individual polyhedral particles relax back to the initial spherical droplets. This is well illustrated by Figure 4b, where the initial and final droplet size distributions are compared. Thus, the HIPE generated by the present process provide an efficient way to reversibly increase viscoelasticity to levels beyond those encountered with standard gels, while preserving their entire processability.

[pic 7]

Figure 4. (a) Dependence between the shear modulus G’ and the radius of oil droplet templates. A G’~1/r scaling law is obtained, the exact prefactor depending on the type of oil used and the protein crosslinking process used (the two set of points correspond to two different oils). (b) Initial (black symbols) and re-hydrated (white symbols) particle size distributions.

4. CONCLUSIONS

We have shown that the main source of viscoelasticity in complex food systems, that is the interfacial tension, can be tuned and designed at different length scales.

At low length scales, liquid crystalline systems based on self-assembled lipid and water, offer a very efficient opportunity to vary the viscoelastic response by changing the exact structure: plastic, viscoelastic and highly elastic fluids can thus be targeted by designing lamellar, hexagonal or bicontinuous cubic phases.

Emulsions, and in particular high internal phase emulsions, also constitute an efficient system in which the viscoelastic response can be controlled by the intrinsic interfacial tension, the interfacial elasticity (for crosslinked interfaces), and the typical size of emulsion droplets.

6. REFERENCES

(1) de Gennes, P.G.; The physics of Liquid Crystals, Oxford University