Shape and Interfacial Viscoelastic Response of Emulsion Droplets in Shear Flow - Food Colloids: Self-Assembly and Material Science - page 349

SDS using oscillating pendant drops did not show any measurable interfacial
tension response, i.e., the interfacial relaxation process is presumed to occur
much faster than our measurements. Static interfacial tensions (c SDS ¼ 1 wt.%,
201C) were 4 mNm 1 in aqueous solution (used for the parent emulsion) and 3.9
mN m 1 when 40wt.% PEG was added.
A typical set of results for the frequency-dependent storage and loss moduli
of the dilute emulsion in the absence of surface rheological effects is shown in
Figure 3. If both the disperse phase and the continuous phase are Newtonian,
we observe a 'relaxation shoulder' in the elastic modulus centred above a
characteristic frequency related to the shape relaxation of the droplets. The
related relaxation timescale is influenced by the interfacial tension, the contin-
uous phase viscosity, the viscosity ratio and the droplet size. For the latter, a
mean radius derived from the droplet-size distribution can be used. Emulsion
rheological models, such as those by Palierne 39 or Yu et al. 40 can be used to
describe emulsion rheological data with shape relaxation. In Figure 3(c)
calculated values for G 0 ( o ) are included (dashed line). The discrepancy above
the relaxation shoulder is due to the use of a mean droplet diameter rather than
the full size distribution. An alternative representation of these data is shown in
Figure 3(d): the relaxation time spectrum l H( l ) of the droplets was calculated
from the complex modulus G*( o ) using standard algorithms. 41 In this repre-
sentation we can directly relate the relaxation timescale, as indicated by the
maximum in the normalized spectrum, to the droplet relaxation process.
We note that, in the absence of dispersed droplets, a zero relaxation time
spectrum is derived, corresponding to Newtonian scaling with G 0 p o 2 and
G 00 p o (see Figure 3(a)).
In Figure 4 the storage moduli of dilute emulsions ( f ¼ 4.5 vol.%) prepared
with excess SDS or b-lactoglobulin are compared. We note the absence of a
characteristic relaxation shoulder for interfaces stabilized by the protein.
The higher interfacial tension of the protein-stabilized system is unlikely to
be responsible for this qualitative difference in the frequency response: if we
compare the characteristic timescales calculated according to the Oldroyd
equation, 22 a mere shift in the characteristic frequency by about a decade
would be observed. If the (static) interfacial tension alone would govern the
stress boundary condition between the oil and water phases, we would expect
comparable shape responses at comparable Capillary numbers. Therefore, the
difference between the surfactant and the protein data appears to be qualitative
in nature: it appears that in the protein system the deformation of the micro-
metre-sized droplets is completely suppressed.
To probe the morphology of the emulsions under flow, we use light-scatter-
ing patterns obtained at the maximum deformation rate during an oscillation
cycle in the rheometer, i.e., when the oscillatory deformation g (t) passes
through zero. In these experiments, the optical technique visualizes the mor-
phology in the velocity vorticity plane (parallel-plate measuring geometry).
For the emulsions stabilized with excess SDS, a characteristic distortion of the
light-scattering patterns in the vorticity direction is observed. Such anisotropic
scattering

patterns have been observed both with phase-separated