Full text of DOI:10.1111/j.1365-2621.2001.tb15592.x

JFS: Food Engineering and Physical Properties
Comparison of the Emulsifying Properties of
Fish Gelatin and Commercial Milk Proteins
E. DICKINSON AND G. LOPEZ
ABSTRACT: We have compared the flocculation, coalescence, and creaming properties of oil-in-water emulsions
prepared with fish gelatin as sole emulsifying agent with those of emulsions prepared with sodium caseinate and
whey protein. Two milk protein samples were selected from 9 commercial protein samples screened in a prelimi-
nary study. Emulsions of 20 vol% n-tetradecane or triglyceride oil were made at pH 6.8 and at different protein/oil
ratios. Changes in droplet-size distribution were determined after storage and centrifugation and after treatment
with excess surfactant.We have demonstrated the superior emulsifying properties of sodium caseinate, the suscep-
tibility of whey protein emulsions to increasing flocculation on storage, and the coalescence of gelatin emulsions
following centrifugation.
Key Words: emulsification, gelatin, caseinate, whey protein isolate, emulsion stability
Introduction
water emulsions of 20% dispersed phase were made under
LTHOUGH THE INGREDIENT GELATINE IS PRIMARILY KNOWN identical homogenization conditions using 2 different oil
A
for its thermoreversible gelation behavior, it has many types—a pure hydrocarbon oil (n-tetradecane) and a com-
other functional applications in food formulations including mercial triglyceride oil (rich in triolein). Observations have
water-holding, thickening, colloid stabilization, crystalliza- been made of the effects of emulsifier origin and concentra-
tion control, film formation, whipping, and emulsification tion on emulsion droplet-size distributions and creaming
(
Ward and Courts 1977; Hudson 1994). The versatility of the stability. Changes in average droplet size with storage time,
protein gelatin as a hydrocolloid is particularly valued in and also following centrifugation and/or surfactant addition,
products like emulsified powders (Kläui and others 1970) have been used to assess the comparative extents of floccu-
where its surface-active and film-forming characteristics can lation and coalescence.
be successfully exploited during the emulsification process
and its stabilization and gelation characteristics during the
subsequent drying and encapsulation stages.
Materials and Methods
HE GELATIN SAMPLE G WAS A COMMERCIAL FISH GELATIN
from Norland Products Ltd (Nova Scotia, Canada). This
The class of food proteins most commonly used for their
T
emulsification properties under neutral pH conditions are low-viscosity gelatin (about 40 kDa) is of a type used by
the milk proteins (Morr 1982; Mulvihill and Fox 1989; Dickin- Hoffmann-La Roche for edible oil emulsification and encap-
son 1997). Two milk protein ingredients widely used for sulation. Unlike most mammalian gelatin samples, it does
emulsification are sodium caseinate and whey protein isolate not form a gel under the standard conditions for determin-
(
WPI). For commercial ingredients that are reasonably pure ing Bloom strength (Leuenberger 1991).
and have not been abused by thermal (or other) processing,
Of the 6 casein samples, C1, C2, and C3 were obtained
it has been reported (Foley and O’Connell 1990) that sodium from New Zealand Milk Products (Rellingen, Germany); C4
caseinate and WPI give similar emulsifying capacity at neutral and C5 from Lactoprot (Kaltenkirchen, Germany); and C6
pH, despite their different adsorbed layer structures (Dickin- from DMV (Veghel, Netherlands). Of the 3 whey protein
son 1997; Dalgleish 1999). Nevertheless, there are some im- samples, W1 was from DMV and W2 and W3 from Davisco
portant functional differences between the caseins and whey (Le Sueur, Minn., U.S.A.). Sample C1 (Alanate 188) was a
proteins, especially relating to their aggregation properties spray-dried sodium caseinate (moisture 3.6%, fat 1.1%, ash
before and after emulsification.
3.6%, calcium 20 mg/100 g), and sample C2 (Alanate 351) was
For certain food emulsion formulations, there is interest in the equivalent potassium caseinate. Sample C3 (Alaco 7005)
replacing gelatin by milk proteins, and vice versa. While it is al- was a casein hydrolysate (moisture 4.5%, fat 0.8%, ash 5.6%,
ready known that gelatin forms rather coarse emulsions (Ches- degree of hydrolysis 33%). Samples C4 (SHV 13) and C5 (LV)
worth and others 1985) due to its lower surface-activity at the were sodium caseinates (5.4% moisture, 1.5% fat) of ash con-
oil–water interface than casein or whey protein (Dickinson and tents 5.6% and 4.2%, respectively. Sample C6 (EM 7) was a
others 1985, 1989), there has been limited direct comparison of sodium caseinate (moisture 5.0%, fat 0.8%, ash 4.0%). Sample
the stability of oil-in-water emulsions made with gelatin and W1 (Esprion 580) was an ultra-filtrated spray-dried whey
milk proteins under the same well-controlled conditions.
protein concentrate (moisture 4.5%, lactose 4.0%, fat 7.5%,
This paper compares the emulsion-stabilizing properties ash 2.5%). Sample W2 (WPI-LE-001-8-919) was a hydrolyzed
of a set of commercial casein and whey protein ingredients whey protein isolate (moisture 4.0%, lactose Ͻ1%, fat Ͻ 1%,
under neutral pH conditions with the properties of a type of ash 5.5%, degree of hydrolysis 6%). Sample W3 (WPI JE 026-
fish gelatin that is currently used industrially as an emulsify- 7-420) was a whey protein isolate (moisture 4.7%, lactose Ͻ
ing agent in oil-soluble vitamin encapsulation. Fine oil-in- 0.5%, fat 0.6%, ash 1.7%).
1
18 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 1, 2001
© 2001 Institute of Food Technologists

Emulsifying Properties of Food Proteins . . .
The n-tetradecane (Ͼ 99%) was obtained from Sigma peak centered around d equals 6 ␮m. Curve (b) is the distri-
Chemicals (St. Louis, Mo., U.S.A.). The commercial triglycer- bution function in the presence of Tween 20. We see that, fol-
ide oil (Trisun 80) was obtained from Danisco (Brabrand, lowing the addition of excess surfactant, the original peak
Denmark); its triglyceride fatty-acid composition was 80% centered around d about 6 ␮m has disappeared and is re-
oleic, 9% linoleic, 4% stearic, and 4% palmitic. Aqueous placed by a skewed monomodal peak centered around d
phosphate buffer solutions (pH 6.8, 0.05 M) were prepared about 0.6 ␮m. It can be inferred that the original major peak
using analytical-grade reagents and double-distilled water. in curve (a) was not due to larger spherical droplets but rath-
Sodium azide (0.1%) was used as antimicrobial agent. The er to aggregates of the 0.6 ␮m primary droplets held togeth-
nonionic surfactant Tween 20 (polyoxyethylene sorbitan er by bridges of protein emulsifying agent. Dissociation of
monolaurate) was used as a dispersing agent.
the flocs occurs on addition of surfactant, and so curve (b)
corresponds to the actual size distribution of the primary
emulsion droplets.
Emulsion preparation and characterization
Oil-in-water emulsions (20 vol% oil, 80% vol% protein so-
Figure 1B shows droplet-size distributions for an equiva-
lution) were prepared at 20 Ϯ 2 ЊC under standard conditions lent low-salt caseinate-stabilized emulsion. The initial bimodal
using a laboratory-scale jet homogenizer (Burgaud and oth- distribution function P(d) has a major peak centered around d
ers 1990) operating at a nominal pressure of 300 bar. The about 0.6 ␮m and a minor peak centered around 4–5 ␮m. In
aqueous phase consisted of a solution of the protein in 0.05 this case, the addition of Tween 20 leads to no significant
M phosphate buffer (pH 6.8) containing 0.1 % sodium azide change in the form of the distribution function. We may there-
(
in some cases 0.5 M salt was also added). Emulsion droplet- fore infer that this emulsion sample is essentially unflocculat-
size distributions were determined by multi-angle static light ed, and that the second peak in curve (a) is really due to the
scattering using a Malvern Mastersizer 2000G (Malvern In- presence of large oil droplets, formed by recoalescence—dur-
struments, Malvern, U.K.) with assumed values of the parti- ing, or shortly after, homogenization—as a result of there be-
cle absorption parameter of 0.007 for n-tetradecane and ing insufficient proteinaceous emulsifier present in relation to
0
.005 for Trisun oil. For the purposes of monitoring stability the new surface area created. Since replacement of adsorbed
and comparing between systems, the average droplet size protein by adsorbed surfactant under quiescent conditions
4
3
was taken as the quantity d43 ϭ ⌺ n d /⌺ n d , where n is does not affect the primary oil droplet distribution, curve (b)
i
i
i
i
i
i
i
the number of droplets of dia di.
in Fig. 1B has a very similar shape to curve (a).
Creaming stability was determined by monitoring the ex-
tent of visible serum separation in emulsion samples of
height 50 mm stored quiescently at 20 ЊC. Stability with re-
spect to flocculation/coalescence was determined by moni-
toring changes in average droplet size d43 with time in regu-
larly mixed emulsion samples. The destabilization was accel-
1A
4
erated by centrifugation of some samples at 2 ϫ 10 g for 30
min and then redetermination of the droplet-size distribu-
tion after careful redispersion of the separated cream layer.
In order to distinguish between flocculation and coales-
cence, emulsion samples were treated with excess surfactant
(
2% Tween 20) immediately after emulsification or following
quiescent storage or centrifugation.
Results and Discussion
Assessing the extent of flocculation
Good emulsification behavior is generally indicated by a
narrow monomodal droplet-size distribution and a small av-
erage size. A broad or bimodal droplet-size distribution, as
determined by multi-angle static light scattering, may be in-
dicative of the presence of a population of larger droplets
caused by inadequate homogenization and/or coalescence
following storage or centrifugation. Alternatively, a bimodal
distribution may indicate the presence of a population of
nonreversibly flocculated droplets formed during or shortly
after homogenization. Distinguishing between large individu-
al protein-coated droplets and flocculated protein-bridged
droplets can be achieved by treating the emulsion with a
low-molecular-weight surfactant. Addition of excess surfac-
tant leads to most of the protein being displaced from the
oil–water interface and to the subsequent stabilization of the
individual oil droplets by the added surfactant.
1B
Fig. 1—Examples of (A) a flocculated emulsion and (B) a
nonflocculated emulsion. The volume-weighted distribu-
tion function P(d) for droplets of dia d is plotted for (A) a
Figure 1A shows droplet-size distributions for a high-salt gelatin-stabilized emulsion (20 vol% n-tetradecane, 0.5 wt%
gelatin (sample G), pH 6.8, 0.5 M NaCl, 20 °C) and (B) a
gelatin-stabilized emulsion of relatively low protein/oil ratio.
Curve (a) represents the initial distribution function P(d) im-
mediately after emulsification. It is strongly bimodal with a emulsification; (b) after emulsification and addition of 2
minor peak centered around d equals 0.6 ␮m and a major wt% Tween 20.
casein-stabilized emulsion (20 vol% n-tetradecane, 0.5 wt%
casein (sample C1), pH 6.8, 20 °C): (a) immediately after
Vol. 66, No. 1, 2001—JOURNAL OF FOOD SCIENCE 119

Emulsifying Properties of Food Proteins . . .
Table 1—Comparison of the stability properties of n-tetradecane oil-in-water emulsions (20 vol% oil, 1 wt % protein,
pH 6.8, 20 °C) made with 10 different protein emulsifiers
a
Serum layer b
thickness (mm)
Protein
emulsifier
Average droplet
Flocculation
stability c
Coalescence
stability
d
size d43 (␮m)
sample
type
sample
code
after
24 h
after
1 wk
after
24 h
after
1 wk
gelatin
casein
casein
casein
casein
casein
casein
whey
G
C1
C2
C3
C4
C5
C6
W1
W2
W3
1.2
0.76
0.72
28
0.78
0.75
0.74
6.4
5.5
0.77
0.77
30
0.77
0.79
0.76
8.8
1
1
1
32
1
1
1
1
15
1
5
2
2
34
2
2
ϫ
ϫ
ϫ
2
13
19
13
ϫ
ϫ
ϫ
ϫ
ϫ
whey
whey
13
1.5
10
3.9
a
Estimated experimental error ± 5%
Total sample height = 50 mm; estimated experimental error ± 1 mm
b
c
d
Key:
Key:
= no flocculation evident; ϫ = flocculation detected
= no coalescence evident; ϫ = coalescence detected
Table 2—Comparison of the stability properties of triglyceride oil-in-water emulsions (20 vol% oil, 1 wt % protein, pH
6
.8, 20 °C) made with 10 different protein emulsifiers
a
Serum layer b
thickness (mm)
Protein
emulsifier
Average droplet
Flocculation
stability c
Coalescence
stability
d
size d43 (␮m)
sample
type
sample
code
after
24 h
after
1 wk
after
24 h
after
1 wk
gelatin
casein
casein
casein
casein
casein
casein
whey
G
C1
C2
C3
C4
C5
C6
W1
W2
W3
7.1
0.74
0.84
57
0.87
0.82
0.83
6.9
21
1
1
1
24
1
1
1
1
10
1
2
2
2
25
2
2
2
2
26
2
ϫ
ϫ
ϫ
ϫ
0.82
0.85
71
0.83
0.83
0.90
10
ϫ
ϫ
whey
whey
7.4
0.78
5.7
0.73
ϫ
a
Estimated experimental error ± 5%
Total sample height = 50 mm; estimated experimental error ± 1 mm
b
c
d
Key:
Key:
= no flocculation evident; ϫ = flocculation detected
= no coalescence evident; ϫ = coalescence detected
The preceding examples refer to 1 emulsion system ex- n-tetradecane-in-water emulsions (d43 Ͻ 0.8 ␮m) with no
hibiting very extensive flocculation (Fig. 1A) and another ex- significant change in average droplet size over the storage
hibiting no significant flocculation (Fig. 1B). Sometimes, period considered. A similar situation arises also with the
however, the behavior can lie intermediate between these ex- caseinate-stabilized triglyceride oil-in-water emulsions (Ta-
tremes. That is, there is an initial bimodal distribution, and ble 2), although the average droplet sizes were slightly larger
on addition of excess surfactant, the average size is shifted to (d43 Յ 0.9 ␮m). There was no visibly apparent (or inferred)
smaller d values, but the distribution still remains bimodal. flocculation/coalescence in any of the casein-stabilized
In such cases, we may infer that the emulsion prior to Tween emulsions (except with sample C3, see below), and the
2
0 addition contains a mixture of some large individual drop- creaming stability was also excellent (Ͻ 5% serum separation
lets and some flocculated droplets.
after 1 wk).
At the other extreme of behavior, the casein hydrolyzate
(sample C3) was found to be a very poor emulsifying agent,
Comparison of the 10 protein emulsifiers
A preliminary comparison of the emulsion-stabilizing leading to highly coarse emulsions, with rapid creaming, and
properties of the 10 protein samples (G, C1-6, W1-3) was car- associated extensive evidence of flocculation and coales-
ried out with a set of systems containing 1 wt% protein. Table cence (although no apparent change in stability between 24 h
1
shows the results for n-tetradecane as the dispersed phase and 1 wk). Also rather poor in terms of the measured stabili-
and Table 2 for triglyceride oil as the dispersed phase.
ty parameters was the whey protein hydrolysate (sample
The gelatin-stabilized emulsions showed considerable W2), with either n-tetradecane or triglyceride as the oil
growth in average droplet size d43 over the storage period. phase.
There was evidence for flocculation with both types of oil
The whey protein sample W3 gave fine triglyceride oil-in-
phase and also some coalescence of triglyceride droplets. water emulsions that were as stable as the caseinate systems.
Nevertheless, the stability with respect to serum separation However, the equivalent n-tetradecane emulsions were sub-
appeared moderately satisfactory, in comparison with the ject to significant flocculation and serum separation after 1
whey protein samples (see below).
wk of storage. The whey protein sample W1 produced coars-
From the values of d43 in Table 1 after 24 h and 1 wk, we er emulsions than sample W3, although the overall stability
can see that all the caseinate samples (C1, C2, C4–6) gave fine was roughly similar.
1
20 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 1, 2001

Emulsifying Properties of Food Proteins . . .
Table 3—Effect of centrifugation (2 ϫ 104 g, 30 min) and
surfactant addition (2 wt% Tween 20) on average droplet
size in n-tetradecane oil-in-water emulsions (20 vol% oil,
Comparison of 3 protein emulsifier class leaders
Based on the results summarized in Tables 1 and 2, proba-
bly the most effective caseinate emulsifier was sample C1—
with others, such as C2, quite close behind in terms of perfor-
mance. The most effective whey protein emulsifier was sam-
ple W3. Together with the gelatin sample, the milk protein
samples C1 and W3 were selected for more detailed study.
Figure 2 shows the effect of protein emulsifier concentra-
tion on the droplet-size distributions of freshly made n-tet-
radecane-in-water emulsions stabilized by (A) gelatin, (B) so-
dium caseinate, and (C) whey protein isolate (WPI). In the
case of gelatin (Fig. 2A), the concentration of approximately
0
.5 wt% protein, pH 6.8, 20 °C) made with 3 types of pro-
tein emulsifier
Protein
Average droplet dia d43 (␮m)
emulsifier
immediately
after
emulsion formation
after
following
following
emulsion centrifugation centrifugation
and
formation
and
surfactant
addition
surfactant
addition
1
wt% protein is close to the optimum concentration for
gelatin (G)
sodium
caseinate (C1) 1.2
3.5
1.9
1.1
0.8
12
1.1
3.6
10
1.1
0.8
producing fine emulsions (d43 ϭ 0.88 ␮m). Halving or dou-
bling the protein/oil ratio led to emulsions with a substan-
tially greater proportion of large droplets. The poorer emul- whey protein
sifying character of the gelatin at higher concentrations may
isolate (W3)
0.9
be a viscosity effect. Sodium caseinate gave increasingly fine
emulsions (d43 0.5 ␮m) as the protein concentration was
increased to 2.5 wt% (Fig. 2B), but with no significant im-
provement above this value. The WPI was found to have an
apparently slightly lower optimum protein concentration of
2
wt% (d43 ϭ 0.56 ␮m).
In order to compare more sensitively the emulsifying effi-
2
A
ciency of the 3 types of proteins, there is benefit in consider-
ing emulsions prepared at constant low protein/oil ratio. Ta-
ble 3 shows changes in the average droplet size of n-tetrade-
cane-in-water emulsions (20 vol% oil, 0.5 wt% protein) fol-
lowing centrifugation and/or surfactant addition. In terms of
initial average size, we see that the relative efficiencies of the
samples lies in the order W3 Ͼ C1 ϾϾ G. The highly floccu-
lated state of the freshly made gelatin-stabilized emulsion
was indicated by the nearly 50% reduction in d43 following
surfactant addition, in contrast to the barely significant
change in d43 for the equivalent caseinate and WPI. Even at
this low protein/oil ratio, the caseinate-based emulsion was
impressively stable toward centrifugation, which is consis-
tent with its good creaming stability (10 mm serum thickness
after 2 wk). In contrast, the relatively coarse gelatin-based
emulsion creamed rapidly (32 mm serum thickness after 4 d)
and was unstable toward droplet coalescence in the centri-
fuge (d43 increased 3-fold). Although the whey protein-coat-
ed droplets were initially rather smaller than the caseinate-
coated droplets, the whey protein emulsion creamed consid-
erably faster on extended storage (16 mm serum thickness
after 1 wk). This can be explained in terms of the flocculation
occurring after emulsion formation through interdroplet dis-
ulfide bonding (McClements and others 1993). The sensitivity
of the whey protein emulsion to flocculation is well illustrat-
ed by the centrifuge test data in Table 3: There was a 4-fold
increase in d43 following centrifugation that was totally re-
versed on addition of excess surfactant.
2
B
2C
In the presence of 0.5 M NaCl (Table 4), each of the pro-
tein types produced n-tetradecane droplets of larger average
size than in the absence of salt. Although each of the emul-
sions was affected by centrifugation, the relative perfor-
mance of the sodium caseinate is distinctly impressive. While
there was evidence for some limited flocculation of caseinate
droplets following centrifugation, the equivalent gelatin sys-
Fig. 2—Droplet-size distributions P(d) immediately after
emulsification for n-tetradecane-in-water emulsions (20
vol% oil, pH 6.8, 20 °C) made with different concentra- tem showed extensive coalescence, and the whey protein
tions of the 3 types of protein emulsifier. (A) Gelatin (sample system showed both flocculation and coalescence. The re-
G): (a) 0.25 wt%, (b) 0.5 wt%, (c) 0.75 wt%, (d) 1.0 wt%, (e)
sults in Table 4 demonstrate the much more effective steric
2
0
.5 wt%, (f) 4 wt%. (B) Sodium caseinate (sample C1): (a)
.25 wt%, (b) 0.5 wt%, (c) 1.0 wt%, (d) 2.0 wt%, (e) 2.5
stabilizing behavior of adsorbed caseinate than adsorbed gel-
atin or whey protein, when compared under interfacial con-
ditions of relatively low surface coverage (that is, low pro-
wt%. (C) Whey protein isolate (sample W3): (a) 0.25 wt%,
(
b) 0.5 wt%, (c) 1.0 wt%, (d) 2.0 wt%.
Vol. 66, No. 1, 2001—JOURNAL OF FOOD SCIENCE 121

Emulsifying Properties of Food Proteins . . .
Table 4—Effect of centrifugation (2 ϫ 104 g, 30 min) and Table 5—Effect of storage and surfactant addition (2 wt%
surfactant addition (2 wt% Tween 20) on average droplet Tween 20) on average droplet size in n-tetradecane oil-in-
size in n-tetradecane oil-in-water emulsions (20 vol% oil, water emulsions (20 vol% oil, pH 6.8, 20 °C) made with the
0
.5 wt% protein, pH 6.8, 20 °C) made with 3 types of pro- “optimum” protein concentration for 3 types of protein
tein emulsifier in presence of 0.5 M sodium chloride
emulsifier
Protein
Average droplet dia d43 (␮m)
Protein
“Optimum”
Average droplet dia d43 (␮m)
emulsifier
emulsifier concentration
wt%)
(
immediately
after
after
following
following
emulsion centrifugation centrifugation
immediately after
following following
emulsion formation
and
surfactant
addition
after
emulsion storage
storage
and
formation
and
emulsion formation
formation
for
surfactant
addition
and
1 wk surfactant
surfactant addition
addition
gelatin (G)
sodium
8.2
15
28
3.4
34
24
2.0
13
gelatin (G)
sodium
caseinate (C1) 2.5
whey protein
1
0.88
0.50
0.56
0.62
nd a
0.48
1.3
0.54
4.2
0.73
nd a
0.56
caseinate (C1) 1.8
whey protein
1.5
1.3
isolate (W3)
15
isolate (W3)
2
a
nd = not determined (due to assumed absence of flocculation)
3
A
tein/oil ratio) and solution conditions where electrostatic
stabilization is unlikely to make a major contribution (that is,
high ionic strength).
Turning now to the triglyceride oil-in-water emulsions,
the comparative trends of behavior are similar to the hydro-
carbon oil-in-water emulsions, although there are some dif-
ferences in the detail. Figure 3 presents sets of droplet-size
distributions for (a) initial emulsions, (b) after surfactant ad-
dition, (c) following centrifugation, and (d) following centrif-
ugation and surfactant addition. The plots in Fig. 3A show
that the 0.5 wt% gelatin emulsion was partially flocculated af-
ter emulsification and exhibited coalescence on centrifuga-
tion. According to Fig. 3B, the 0.5 wt% caseinate emulsion
was not flocculated before or after centrifugation and was
also stable to coalescence. (Caseinate emulsions can be sus-
ceptible to bridging flocculation and coalescence at much
lower protein/oil ratios (Dickinson and others 1997).) Figure
3
B
3
C shows that the 0.5 wt% whey protein emulsion was rela-
tively free from flocculation in its freshly prepared state;
subsequent centrifugation led to extensive flocculation but
apparently no coalescence.
Finally, let us consider the relative performances of the 3
protein types at their “optimum” concentrations (Table 5). All
3
emulsifiers produced emulsions with submicron-sized
droplets. In terms of smallness of initial average droplet dia,
and also the maintenance of that dia during storage, the so-
dium caseinate can be regarded as the most effective emulsi-
fier. Both the gelatin and (especially) the WPI showed evi-
dence of flocculation in the freshly made emulsion and (es-
pecially) in the emulsion stored for 1 wk. With none of the
emulsifiers was there any indication of significant coales-
cence during storage. While the WPI appears nearly as effec-
tive as the caseinate in terms of initial droplet size, the aggre-
gation of whey-protein-coated droplets during storage clear-
ly has a detrimental effect on the apparent droplet-size dis-
tribution and on the associated creaming instability.
3C
Fig. 3—Effect of centrifugation (2 X 104 g, 30 min) and sur-
factant addition (2 wt% Tween 20) on droplet-size distribu-
tions P(d) of triglyceride oil-in-water emulsions (20 vol% oil,
pH 6.8, 20 °C) for the 3 types of protein emulsifier: (A) gelatin
Conclusions
HIS STUDY AT NEUTRAL PH HAS SHOWN THAT WHERE FISH
T
gelatin is intended as a replacement for milk protein (es-
(
sample G), (B) sodium caseinate (sample C1), (C) whey pro-
pecially caseinate), there is benefit in optimizing the protein/
oil ratio in order to avoid the presence of large droplets,
which may be susceptible to coalescence, especially at high
tein isolate (sample W3). Key: (a) after emulsification; (b)
after emulsification and surfactant addition; (c) following
centrifugation and redispersion in phosphate buffer; (d) fol-
lowing centrifugation, redispersion, and surfactant addition. ionic strength. Conversely, where milk protein is intended as
1
22 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 1, 2001

Emulsifying Properties of Food Proteins . . .
and whey protein isolate in 18% oil in aqueous systems. J Dairy Res 57:377–391.
Hudson CB. 1994. Gelatine: relating structure and chemistry to functionality. In:
Nishinari K, Doi E, editors. Food hydrocolloids: Structures, properties and func-
tions. New York: Plenum Press. P 347–354.
Kläui HM, Hausheer W, Huschke G. 1970. Technological aspects of the use of fat-
soluble vitamins and carotenoids and of the development of stabilized mar-
ketable forms. In: Morton RA, editor. Fat-soluble vitamins, international en-
cyclopedia of food and nutrition. Vol. 9. Oxford: Pergaman.
a replacement for gelatin in existing emulsion products, at-
tention should be given to the effect of flocculation of whey-
protein-coated droplets on storage.
References
Burgaud I, Dickinson E, Nelson PV. 1990. An improved high-pressure homoge-
nizer for making fine emulsions on a small scale. Int J Food Sci Technol 25:39– Leuenberger BH. 1991. Investigation of viscosity and gelation properties of
6.
different mammalian and fish gelatins. Food Hydrocoll 5:353–361.
Chesworth SM, Dickinson E, Searle A, Stainsby G. 1985. Properties of oil-in- McClements DJ, Monahan FJ, Kinsella JE. 1993. Disulfide bond formation affects
4
water emulsions containing gelatin and caseinate. Lebensm Wiss Technol
8:230–232.
Dalgleish DG. 1999. Interfacial structures and colloidal interactions in protein-
stabilized emulsions. In: Dickinson E, Rodriguez Patino JM, editors. Food emul-
sions and foams: Interfaces, interactions and stability. Cambridge, U.K.: Royal Mulvihill DM, Fox PF. 1989. Physicochemical and functional properties of milk
stability of whey protein isolate emulsions. J Food Sci 58:1036–1039.
Morr CV. 1982. Functional properties of milk proteins and their use as functional
ingredients. In: Fox PF, editor. Developments in dairy chemistry—1. London:
Applied Science. P 375–399.
1
Society of Chemistry. P 1–16.
Dickinson E. 1997. Properties of emulsions stabilized with milk proteins. J Dairy
proteins. In: Fox PF, editor. Developments in dairy chemistry—4. London: Elsevi-
er Applied Science. P 131–172.
Sci 80:2607–2619.
Ward AG, Courts A, editors. 1997. The science and technology of gelatin. London:
Academic Press.
Dickinson E, Pogson DJ, Robson EW, Stainsby G. 1985. Time-dependent surface
pressures of adsorbed films of casein + gelatin at the oil–water interface. MS 20000129
Colloids Surf 14:135–141.
Dickinson E, Flint FO, Hunt JA. 1989. Bridging flocculation in binary protein-
stabilized emulsions. Food Hydrocoll 3:389–397.
We thank Hoffmann-La Roche, Vitamins Division (Basel, Switzerland) for the supply of
commercial protein samples and for financial support of this project.
Dickinson E, Golding M, Povey MJW. 1997. Creaming and flocculation in oil-
in-water emulsions containing sodium caseinate. J Colloid Interface Sci
Authors are with the Procter Department of Food Science, University of Leeds,
Leeds LS2 9JT, UK. Direct correspondence to Eric Dickinson (E-mail:
E.Dickinson@leeds.ac.uk).
1
85:515–529.
Foley J, O’Connell C. 1990. Comparative emulsifying properties of sodium caseinate
Vol. 66, No. 1, 2001—JOURNAL OF FOOD SCIENCE 123