,4. 80 60 1 . 0 Q
I
S 20I
A S * •\*5> 0.2 0.4 0.6 0.8D isp ersed Phase Concentration (-)
A -450rpm # -600rpm □ -750rpm a -900rp m
Fig. 5.9. Correlation of Sauter mean diam eter, with Xylene Dispersed Phase Concentration, (j), at Im peller diam eter, Di = 0.05m.
240 1 E 200 -- 3 s 160 -- s E .2 120 - - I . . - u
I
40 -- 0.6 0 0.2 0.4 0.8 1 D ispersed PhaseConcentration (-) A -200rpm ♦ -300rpm □ -550rpm ♦ -700rpmFig. 5.10. Correlation of Sauter mean diam eter, d]2 with Sunflower oil Dispersed Phase C oncentration, ({>, at Im peller diam eter, D, = 0.05m.
Chavter 5_________ ;____________________________________ Results and Discussion
The shift in size distribution to smaller average drop sizes for (j) > 50% was not clearly
distinguishable (cf. Fig. 5.8b and 5.8c). Figures 5.9 and 5.10 are a simplified
representation of Figs. 5.7 and 5.8 respectively. In these plots the different markers
represent the change in Sauter mean diameter with increasing dispersed phase
concentration for the four impeller speeds considered for each liquid dispersion system.
From the drop size distribution (DSD) plots it can be deduced that the average drop
size increases with an increase in dispersed phase concentration, ({) up to a point beyond
which further increase in the dispersed phase concentration causes the average drop size
to decrease. Although the results for sunflower oil are not clearly evident, these results
are in general agreement with the data reported recently by Boye et al, (1996) and
previously by Clarke and Sawistowski, (1978), Kumar et al, (1991) and Boye and
Ayazi Shamlou, (1994). The dashed lines represent the theoretical predictions of the
model postulated and once more seem to support these findings.
Experiments were conducted for aU three impellers studied and the results are illustrated
in appendix (C3.3/C4.3). In cases where there was excessive vortexing and air
entrainment the experiment was aborted (especially Di = 0.08 m). Clarke and
Sawistowski (1978) suggested that the observed change in the dependence of drop size
on the dispersed phase concentration was due to phase inversion which generally occurs
when the dispersed drops are packed closely in the suspension. Following phase
inversion the continuous phase becomes the dispersed phase and the dispersed phase
substituted by the continuous phase. Phase inversion is usually characterised by a step
change in some of the key physical properties of the two-liquid phase mixture, for
Chapter 5_____________________________________________ Results and Discussion
(Kato et a l, 1991). The type of emulsifier used in a dispersion has profound influence
on its stabilising or destabilising properties in addition to the hydrodynamic conditions in
the stirred vessel which ultimately determine the structure of the dispersion.
Matsumoto (1983) has described a method of producing duplex emulsion and dispersion
structures through thermal induction by assessing the extent of oil layer on the surface of
the dispersed aqueous compartments using a viscometric method. This method requires
measuring viscosity changes under glucose concentration gradients between inner
aqueous suspending fluid in order to estimate the area of the layer separating the two
aqueous phases. He also observed that the area of layer separating the two aqueous
phases mainly depended on the molar ratio of the polymer used (e.g. polyoxyethylene
sorbitan monooleate - Span 80) to the hydrophilic emulsifier in the sample.
Anionic surfactants such as sodium dodecyl sulphate (SDS) used in this study appear to
provide a wider molar ratio (4 60) than non-ionic polymers (polyoxyethylene sorbitan
monooleate - Span 80); cetyl trimethyl ammonium bromide (CTABr)) for the formation
of complex water-oil-water type (W/OAV) dispersions. To determine the formation of
global phase inversion, other experimental measurements such as the dispersion
electrical conductivity and rheological behaviour are considered in the subsequent
Chapter 5______________________________________________Results and Discussion
5.2.1 EFFECT OF DISPERSED PHASE C O NCENTRATIO N ON DISPER SIO N ELECTRICA L CO NDUCTIVITY
The data in Figs. 5.11a and 5.11b represent typical electrical conductivity measurements
and refer to experiments which were carried out at a fixed speed of 450 rpm for xylene
dispersions and 200 rpm for sunflower oil dispersions respectively. For both two-phase
liquid dispersions, several measurements of electrical conductivity were made for each
dispersed phase concentration as a function of time shown in both plots and for
comparison the conductivity of the pure dispersed phase and aqueous phase containing 0.3
wt% SDS are also depicted. The results show a gradual change in the electrical
conductivity of the dispersion with increasing dispersed phase concentration, but there is
no evidence of a sharp change in the conductivity that can be associated with the
phenomenon of global phase inversion (appendix C3.4/C4.4).
These results are contrary to those reported by Quinn and Sigloh (1963) and more
recently by Brooks and Richmond (1991) who observed a sharp change in electrical
conductivity. Moreover, visual assessment of the vessel contents showed no significant
change in the colour of the dispersion as dispersed phase concentration for both systems
increased. In most of the experiments, the mixture had a shiny appearance, characteristic
of an aqueous-continuous phase dispersion. In a few cases however, especially those
involving very high dispersed phase concentrations and speed of agitation, the dispersion
became dull in its appearance. In these cases, it was not obvious whether the change in
colour had occurred as a result of phase inversion or émulsification. The results from these
experiments were excluded from further analysis. Additional experiments such as
rheology measurements and visual assessment of the dispersion were performed to
Chapter 5 Results and Discussion