Food Chemistry 309 (2020) 125758

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Interfacial and emulsification properties of sono-emulsified grape seed oil T

emulsions stabilized with milk proteins
⁎
Mayumi Silvaa, Bogdan Zisub, Jayani Chandrapalaa,
a
School of Science, RMIT University, Bundoora, VIC 3083, Australia
b
Spraying Systems, Fluid Air, Melbourne, Australia

A R T I C LE I N FO A B S T R A C T

Keywords: Emulsions were designed under low frequency ultrasound (20 kHz) at energy densities of 11.7–117.0 J/mL using
Sono-emulsion grape seed oil and milk protein solutions containing different casein to whey protein ratios of 80:20, 60:40,
Milk proteins 50:50 and 40:60. An increase in energy densities produced emulsions with a smaller droplet size and narrow size
Interfacial tension distribution at all milk protein ratios. However, the minimum sono-energy density required to produce stable
Emulsifying ability
emulsions varied depending on the ratio of caseins (CN) and whey proteins (WP) in the continuous phase. In
Grape seed oil
addition, the composition of the interfacial layer was dependent on the composition of the milk proteins in the
continuous phase. The interfacial layer was predominantly covered by the CN and CN-WP aggregates in the
presence of equal or greater amounts of caseins than whey proteins (80:20, 60:40 and 50:50), while WP ag-
gregates and CN-WP aggregates were the primary constituents of whey protein-rich emulsions (40:60).

1. Introduction crosslinking. Milk protein solutions consisting of mostly whey proteins

Caseins and whey proteins have excellent but different emulsifying
characteristics due to the differences in their molecular structures
(Parkinson & Dickinson, 2004). Ultrasound is used as an emulsification
technique in order to improve the emulsifying properties of milk pro-
tein-based emulsions (O'Sullivan, Arellano, Pichot, & Norton, 2014;
Shanmugam & Ashokkumar, 2014). However, low frequency ultra-
sound influences the physicochemical, thermal, and structural proper-
ties of milk protein solutions depending on the protein composition and
ultrasonic conditions in use (Silva, Zisu, & Chandrapala, 2018). Re-
search has been carried out to understand the effect of sonication on
skim milk (Shanmugam, Chandrapala, & Ashokkumar, 2012), micellar
caseins (Chandrapala, Martin, Kentish, & Ashokkumar, 2014;
Chandrapala, Martin, Zisu, Kentish, & Ashokkumar, 2012), whey pro-
teins (Chandrapala et al., 2012; Chandrapala, Zisu, Palmer, Kentish, &
Ashokkumar, 2011) and mixtures of caseins and whey proteins (Silva
et al., 2018). A previous study conducted by Silva et al. (2018) showed
that the sonication-induced physicochemical and structural modifica-
tions of milk proteins highly depended on the ratio of caseins and whey
proteins present in the milk solutions. The application of high intensity
low frequency ultrasound induces the dissociation and denaturation of
β-lactoglobulin while unfolding the dimeric form of β-lactoglobulin
into random coils leading to aggregation via β-sheet intermolecular
produced large amounts of disulphide linked aggregates while the
casein-rich systems produced more hydrophobically mediated ag-
gregates. These composition-dependent structural changes in milk
proteins may influence the emulsifying characteristics.
Emulsifying properties of caseins and whey proteins have been
widely studied individually in terms of processing conditions and pro-
cessing techniques (Dickinson & Golding, 1998; Pichot, 2012;
Ranadheera, Liyanaarachchi, Chandrapala, Dissanayake, & Vasiljevic,
2016). A previous study showed that the emulsifying properties of
casein and whey protein mixtures strongly depended on the ratio of
caseins and whey proteins present, initial protein concentration and
temperature during emulsification by high pressure homogenization
(Surel et al., 2014). Moreover, the composition of the interfacial layer
and the adsorption behaviour of proteins at the oil-water interface are
also influenced by the state of the proteins in the bulk solution
(Dalgleish, 1995; Dickinson, 2001; Dickinson & McClements, 1995).
However, almost no studies have been carried out to understand the
interfacial and emulsifying behaviour of milk systems with varying
casein to whey protein ratios during ultrasonication.
Therefore, a fundamental understanding of sonication-induced in-
terfacial and emulsifying behaviour of milk protein systems containing
varying levels of caseins and whey proteins will be beneficial to im-
prove the physicochemical, functional and storage properties of the

⁎
Corresponding author.
E-mail address: jayani.chandrapala@rmit.edu.au (J. Chandrapala).

https://doi.org/10.1016/j.foodchem.2019.125758
Received 11 August 2019; Received in revised form 14 October 2019; Accepted 20 October 2019
Available online 25 October 2019
0308-8146/ © 2019 Elsevier Ltd. All rights reserved.
M. Silva, et al.
respective emulsions containing GSO. Apart from the emulsification
technology and emulsifying agents, the oil phase plays an important
role on the physicochemical, nutritional and structural aspects of the
food emulsions (McClements, 2015b). Grape seed oil (GSO) which is a
by-product of wine manufacturing, contains high amount (85–90%) of
polyunsaturated fatty acids (PUFA), mainly linoleic acid, tocopherols
and tocotrienols which are isomers of vitamin E (Fernandes, Casal,
Cruz, Pereira, & Ramalhosa, 2013; Madawala, Kochhar, & Dutta, 2012).
The present study examined the effect of low-frequency ultrasound
(20 kHz) on the interfacial and emulsifying properties of milk protein
(3.3% w/v) systems with varying casein to whey protein ratios of 80:20,
60:40, 50:50, 40:60, pure milk protein isolate (MPI) and pure whey
proteins isolate (WPI) during the processing of primary emulsions
containing 10% (w/w) grape seed oil.
2. Materials and methods
2.1. Materials
Milk Protein Isolate (MPI) and Whey Protein Isolate (WPI) powders
were purchased from Tatura milk industries limited (Victoria,
Australia). The MPI and WPI powders contained a total protein content
of 85.5% (w/w) and 88.8% (w/w) respectively. Pure grape seed oil
(GSO) (Azalea Brand, NSW, Australia) was purchased from a local su-
permarket. All the chemicals and filters used were obtained from Sigma
Aldrich Pty Ltd (Castle Hill, NSW, Australia) or Bio-Rad Laboratories
Pty Ltd (Gladesville, NSW, Australia). Ultra-pure (MilliQ) water was
used in all experiments.
2.2. Charcoal treatment for GSO
The interfacial tension of the emulsions depends on the oil prop-
erties such as the origin, purity, polarity of the major lipid molecules
(e.g., triglycerols or terpenes) and the presence of surface active com-
ponents like free fatty acids, monoacylglycerols, diacylglycerol or
phospholipids (McClements, 2015b). These surface active substances
may interfere with the adsorption behaviour and will result in obtaining
inaccurate and inconsistent data (Dopierala et al., 2011). Thus, GSO
was purified using an activated charcoal treatment to remove the sur-
face active impurities. For that purpose, GSO was mixed with grounded
charcoal in 10:1 (w/w) ratio and stirred overnight using a magnetic
stirrer. Then the oil was centrifuged (Beckman Coulter, United States)
at 20 °C and the supernatant (oil portion) was filtered through a
Whatman No.40 filter paper and a 0.45 µm PTFE syringe filter (Acro-
disc® syringe filters, Sigma Aldrich Pty Ltd, NSW, Australia). The
treatment was carried out three times and ensured the absence of sur-
face active materials using DIFT data.
2.3. Methods
2.3.1. Preparation of the aqueous phase
Milk protein solutions with different casein to whey protein (C:W)
ratios of 80:20, 60:40, 50:50 and 40:60 were prepared by dissolving the
required amount of MPI and/or WPI powders in MilliQ water at 40 °C
while keeping the total protein concentration at 3.3% (w/v). Pure MPI
and WPI solutions were used as controls. Initially the required amount
of MPI was added to pre-heated (40 °C) MilliQ water and stirred for
15 min. Then, the WPI powder was added and stirred for another
45 min to ensure complete mixing. The solution was kept overnight at
4 °C for further hydration. On the following day, the pH of the solutions
was adjusted to 6.7 using 1 M HCl and 1 M NaOH.
2.3.2. Preparation of emulsions
Oil-in-water primary emulsions were prepared with grape seed oil
(GSO) and milk protein solutions with an oil volume fraction of 10%
(w/w). Fifty millilitre aliquots of samples were prepared by adding
2
Food Chemistry 309 (2020) 125758
5 mL of GSO into 45 mL of milk protein solution. Then the two phases
were emulsified using a 20 kHz low frequency ultrasonic unit with a
500 W ultrasonic horn (Q Sonica, Bandelin electronic, Berlin, Germany)
at 60% amplitude for different time periods of 1, 3, 5, 7 and 10 min. The
ultrasonic horn tip was positioned approximately 5 mm from the sur-
face of the sample near the oil and water interface. The temperature of
the solutions was maintained at 20 ± 2 °C using an ice bath. The ap-
plied energy densities were 11.7 J/mL, 35.1 J/mL, 58.5 J/mL, 81.9 J/
mL and 117.0 J/mL for time periods of 1, 3, 5, 7 and 10 respectively.
2.3.3. Determination of dynamic interfacial tension (DIFT)
The effect of sonication on the DIFT at GSO-milk interface were
determined with a drop profile tensiometer (PAT-1TM, Sinterface
Technologies, Germany) using the pendent drop method. In order to
obtain sonicated milk systems, 50 mL aliquots of milk protein solutions
were sonicated using a 20 kHz high frequency ultrasonic unit (Q Sonica,
Bandelin electronic, Berlin, Germany) at 60% amplitude for 1, 3, 5, 7
and 10 min. The drop volume was kept constant at 20 µL and the
measurements were conducted for 200 s. The temperature of the mea-
suring cell chamber was maintained at 25 °C using a water bath (Julabo,
Germany). The calibration of the tensiometer was conducted with a
sphere (1.5 mm of radius) at 25 °C. The density of the GSO and protein
solutions were measured using a densitometer (METTLER TOLEDO,
PortableLabTM) at 25 °C. According to our experimental results, the
equilibrium interfacial tension at grape seed oil-water interface was
20.08 ± 0.75 mN/m at 25 °C.
2.3.4. Determination of the percentage of adsorbed proteins, interfacial
protein concentration and surface protein composition
The adsorbed protein percentage (AP%) and interfacial protein
concentration (Г) were determined using a method described Ye
(2008). Emulsions were centrifuged (Beckman Coulter, United States)
at 15,000g for 40 min at 20 °C. After centrifugation, the oil droplets and
adsorbed proteins were separated as a cream layer at the top of the
centrifuge tube and the aqueous phase was settled to the bottom. Then
the aqueous layer was filtered through a 0.22 µm filter (Sigma Aldrich
Pty Ltd, NSW, Australia). The cream layer was dispersed in MilliQ water
and re-centrifuged at 15,000g for 40 min to obtain the washed cream.
The protein contents in aqueous phases of both centrifuged emulsions
and protein solutions were determined by Bradford assay. In brief, the
aqueous phases were diluted 40 fold with MilliQ water and 20 µL of the
diluted aqueous phase was mixed with 1 mL of Bradford reagent (Bio-
Rad Laboratories Pty Ltd, NSW, Australia). Then the mixture was in-
cubated for 5 min at room temperature and the absorbance was de-
termined at 595 nm using a UV–VIS Spectrophotometer (PerkinElmer,
Lambda 35, Australia). The standard curve (R2 = 0.9972) was devel-
oped using Bovine serum albumin (BSA) in the concentration range of
250–1250 µgmL−1. The adsorbed protein content (AP%) and interfacial
protein concentrations (Г, expressed in mg/m2) were determined using
following equations (Eqs. (1) and (2)):
(Cs − Cf ) × 100
AP% =
C0 (1)
where, Cs is the protein concentration in the supernatant of initial milk
protein solution, Cf is the protein concentration in the filtrate of
emulsion, C0 is the initial protein concentration of the milk protein
solution applied for the emulsion preparation (3.3 g/100 mL).
(Cs − Cf ) × D(3, 2)
Γ=
6ϕ (2)
where, D(3,2) is the surface-average diameter of the emulsion obtained
from the Mastersizer and ϕ is the oil volume fraction of the emulsion.
The composition of the proteins at the adsorbed layer was de-
termined using sodium dodecyl sulphate polyacrylamide gel electro-
phoresis (SDS-PAGE) under reducing conditions as described by Ye and
Singh (2000). Briefly, the cream was mixed with SDS sample buffer and
M. Silva, et al. Food Chemistry 309 (2020) 125758

β-Mercaptoethanol followed by heating at 95 °C for 5 min in a boiling

water bath. In order to remove the fat, heated samples were centrifuged
at 2500g for 30 min. Ten microliter aliquots of prepared samples were
injected to precast 4–20% mini-protean gels (Bio-Rad Laboratories Pty
Ltd, NSW, Australia) and the gels were run for 45 min at 200 V. Casein,
whey protein and molecular weight standards (10–250 kDa) were run
simultaneously. Gels were then rinsed with MilliQ water, stained with
Coomasie stain and destained using MilliQ water overnight as described
by Silva et al. (2018). Later, the gels were scanned and the intensities of
the protein bands were determined using ImageJ 1.52e software
(Wayne Rasband, USA). The percentage composition of whey proteins
and caseins were determined as a fraction of the sum total.

2.3.5. Emulsion activity and stability indexes

The emulsion activity index (EAI) and emulsion stability index (ESI)
of the emulsions were measured by a turbidimetric method as described
by Pearce and Kinsella (1978). Immediately after sonication, the
emulsions were diluted 100 folds with 0.1% (w/v) sodium dodecyl
sulphate (SDS) and vortexed for 10 s. The absorbance of the diluted
emulsion was measured using a UV–VIS Spectrophotometer at 500 nm
wavelength. The emulsion activity index (EAI, expressed in m2g−1) was
calculated using the following equation (Eq. (3)):
2 × 2.303 × A 0 × D
EAI =
C × ϕ × L × 1000 (3)

where, A0 is the initial absorbance, D is the dilution factor (100), C is

the concentration of protein in the milk protein solutions (33 mgmL−1),
Φ is the oil volume fraction and L is the light path length (0.01 m).
The emulsions were stored at 4 °C and absorbance was re-measured
after 24 h following the same procedure. The emulsion stability index
(ESI, expressed in hours) was calculated using the following equation
(Eq. (4)):
A × Δt
ESI =
ΔA (4)

where, A is the absorbance after 24 h, Δt is the time interval (24 h) and

ΔA is the change in absorbance during 24 h (A0 − A).

2.3.6. Size of emulsion

Size of the emulsion globules were determined by laser diffraction
method using a Mastersizer 3000 attached to a hydroMV sample
handling unit (Malvern Instruments, Worcester, UK). Emulsions were
directly added to the dispersion unit (containing water) until 15–20%
obscuration level was reached and the measurements were taken at a
rotating speed of 1250 rpm at 25 °C. The refractive indexes of GSO and
milk protein solutions used were 1.471 and 1.336 respectively. The D
(4,3), D(3,2), D10, D50 and D90 values were recorded as an average of
five measurements.
2.3.7. Surface charge of emulsion
Emulsions were diluted 200 times using MilliQ water and the zeta
potential measurements were obtained using the Zetasizer (Nano series,
Malvern Instruments, Worcester, UK) at 25 °C. The refractive indexes
for GSO and milk protein solutions were considered as 1.471 and 1.336
respectively. Three replicate measurements were made for each sample.
2.3.8. Statistical analysis
All the experiments were at least duplicated and reported as the
average with standard deviation. One-way ANOVA was used to de-
termine the significant differences between and within the experiments
at 95% confident level. Pearson correlation coefficient (r) was calcu-
lated to analyse the relationship between emulsifying and physico-
chemical properties. All the statistical analysis was performed with
SPSS version 25 software (SPSS, IBM).
Fig. 1. Dynamic interfacial tension (DIFT) behaviour at oil/water interface
between purified grape seed oil and milk protein solutions (casein:whey −
80:20, 60:40, 50:50, 40:60, MPI and WPI) sonicated at different energy den-
sities (A-untreated, B-11.7 J/mL and C-117.0 J/mL).
3. Results
3.1. Dynamic interfacial tension (DIFT)
The ease of the interfacial layer formation between the oil phase and
the aqueous phase is denoted by the dynamic interfacial tension (DIFT)
at oil–water interface (McClements, 2015a). As shown in Fig. 1, the
DIFT between purified GSO and sonicated milk protein solutions pre-
pared with different caseins to whey protein (C:W) ratios of 60:40,
50:50 and 40:60 were independent from the applied energy density.
However, DIFT at C:W ratio of 80:20 was slightly increased at 11.7 J/
mL energy density compared to the untreated solution followed by a
slight reduction on prolong sonication (117.0 J/mL) simultaneously
showing the highest DIFT among all ratios under all energy densities. In
addition, the decrease in DIFT was observed for sonicated MPI and WPI.
Interestingly, milk protein solutions stabilized with MPI and other ra-
tios attained its equilibrium interfacial tension value immediately after
the oil drop formation (~20 s) in comparison to WPI which took a re-
latively longer time (~180 s) for attaining the equilibration.

3
M. Silva, et al. Food Chemistry 309 (2020) 125758

Table 1
The percentage of adsorbed protein, surface load, emulsion activity index and emulsion stability index of GSO-M emulsions prepared using different casein to whey
protein ratios of 80:20, 60:40, 50:50, 40:60, MPI and WPI under different energy densities of 11.7 J/mL, 35.1 J/mL, 58.5 J/mL, 81.9 J/mL and 117.0 J/mL.
Energy density (J/mL) Adsorbed protein (%) Surface load (mg/m2)

80:20 60:40 50:50 40:60 MPI WPI 80:20 60:40 50:50 40:60 MPI WPI

11.7 2.48a 3.70a 8.62a 4.44a 2.42a 3.77a 0.32a 1.88a 0.80a 2.84a 0.43a 1.56a
35.1 5.49b 6.67b 10.17a 6.20b 5.72b 6.60b 1.55b 2.65b 4.25b 2.76a 1.66b 2.28b
58.5 6.63c 9.90c 14.55b 8.96c 8.48c 16.36c 2.31c 3.45c 5.04c 2.86a 2.46c 4.90c
81.9 8.05d 12.12d 20.81c 12.59d 9.56d 16.77c 2.67c 3.29c 6.41d 3.68b 2.76d 4.64c
117.0 13.64e 22.36e 26.06d 13.67e 14.13e 26.20d 2.70c 5.74d 7.68e 3.66b 2.88d 6.56d
EAI (m2/g) ESI (h)
11.7 15.85a 17.94a 1.17a 18.04a 12.54a 30.68a 88.16a 45.18a – 38.42a 26.57a 170.93a
35.1 20.48b 29.55b 9.10b 33.60b 26.72b 32.91b 103.44b 221.05b 128.49a 209.28b 73.63b 429.95b
58.5 34.85c 34.23c 29.56c 34.47b 33.80c 34.84c 754.29c 833.13c 171.98b 380.99c 727.55c 1609.96c
81.9 35.69d 34.08c 33.80d 34.89b 34.69d 34.25c 805.41d 1716.83e 843.34d 817.12d 948.52d 2104.77e
117.0 36.32e 35.20c 34.98e 34.00b 31.55d 34.59c 1295.32e 1211.27d 529.55c 1356.85e 729.56c 1760.40d

Values in the same column and row with different capital and simple superscripts respectively are significantly different (p < 0.05).

3.2. Percentage of adsorbed proteins, interfacial protein concentration and

surface protein composition

The adsorbed protein percentage (AP%) and the surface protein

concentration of emulsions were largely affected by the casein to whey
protein ratios and the applied sono-energy density (Table 1). The AP%
and surface load increased (p < 0.05) with increasing energy density
while the highest raise in AP% was observed at low energy densities
with the presence of whey proteins in the emulsions. For instance, the
highest raise in AP% was observed in WPI solutions under 58.5 J/mL
energy density while C:W ratios of 80:20 and 60:40 showed the highest
raise in AP% under 117. J/mL energy density. Moreover, emulsions
stabilized with C:W ratio of 50:50 exhibited the highest AP% under all
energy densities and the highest surface load was also observed in the
same system under all energy densities except 11.7 J/mL. However,
under the 117.0 J/mL energy density, ~57% caseins and 43% whey
proteins were presented in the interfacial layer of C:W ratio of 50:50
(Fig. 2A). Furthermore, the lowest AP% and surface load resulted in
emulsions prepared by C:W ratio of 80:20 under all sonication times
with the maximum adsorption of ~14% and 2.70 mg/m2 of surface load
(Table 1) under 117.0 J/mL energy density. Moreover, the interfacial
layer in the same system primarily consisted of casein proteins denoting
~62% fraction (Fig. 2A).

Fig. 2. A – Relative proportion of adsorbed caseins and whey proteins at the

3.3. Size of emulsions
Upon sonication, the width of the bimodal particle size distribution
decreased significantly, creating a unimodal distribution with smaller
emulsion droplets in all milk protein ratios depending on the compo-
sition of milk proteins in the aqueous phase of emulsion (Fig. 3A).
Emulsions having a higher whey protein fraction obtained unimodal
size distribution within shorter sonication times. For instance, emul-
sions stabilized by WPI and C:W ratio of 40:60 attained unimodal dis-
tribution below 58.5 J/mL energy density whereas 81.9 J/mL was re-
quired for C:W ratio of 60:40 and 50:50. However, caseins-rich systems
(MPI and 80:20) attained unimodal distribution up on 117.0 J/mL en-
ergy density.
Droplet size decreased (p < 0.05) with increase in the applied
energy density from 11.7 J/mL to 117.0 J/mL regardless of the com-
position of the proteins in the system (Fig. 4A). Emulsions prepared
with C:W ratio of 40:60 showed the highest particle size (3.13 µm)
whereas smallest particle size (1.58 µm) was observed in WPI emulsions
under 1 min sonication. However, all the emulsions acquired sig-
nificantly similar droplet size (p > 0.05) at higher applied energy
density (117.0 J/mL). The highest percentage of reduction in particle
size was observed at 35.1 J/mL energy density followed by a gradual
decrease in size thereafter at higher sono-energy. Moreover, the
droplet surface (cream phase) of GSO-M emulsions, B – photographs of emul-
sions; 80:20, 60:40, 50:50 (top – left to right), 40:60, MPI and WPI (bottom –
left to right) at different applied energy densities; 11.7 J/mL (1), 35.1 J/mL (2),
58.5 J/mL (3), 81.9 J/mL (4), 117.0 J/mL (5) and protein solutions (P). Layer
separation can be observed in following emulsions; 80:20 – (1), 60:40 – (1 & 2),
50:50 - (1, 2 & 3), 40:60 – (1) and MPI – (1).
decrease in droplets size as the applied energy density increases re-
sulted in adsorption of more proteins at the oil-water interface
(Fig. 3B).
3.4. Emulsion activity and stability indexes
The ability of proteins to form an emulsion is described by the
emulsion activity index (EAI) (Pearce & Kinsella, 1978) and the stability
of the emulsions over a defined time period is described by the emul-
sion stability index (ESI). The EAI and ESI depend on the milk protein
ratio in the aqueous phase and applied energy density (Table 1).
Emulsions stabilized with WPI showed the highest initial EAI of
~30 m2/g while the lowest EAI (~1 m2/g) was observed in C:W ratio of
50:50 at 11.7 J/mL energy density. However, all the emulsions ex-
hibited significantly higher EAI (> 30 m2/g) values at higher applied

4
M. Silva, et al. Food Chemistry 309 (2020) 125758

Fig. 3. A – Volume size distribution, B – relationship between particle size and percentage of adsorbed proteins in GSO-M emulsions prepared with varying casein to
whey ratios of 80:20, 60:40, 50:50, 40:60 and WPI and MPI at different energy densities of 11.7 J/mL, 35.1 J/mL, 58.5 J/mL, 81.9 J/mL and 117.0 J/mL. Capital and
simple superscripts denote the significant difference between sonication times and milk protein ratios respectively.

5
M. Silva, et al.
Fig. 4. A – Volume mean diameter D(4,3), B – surface charge of GSO-M
emulsions prepared using different casein to whey protein ratios of 80:20,
60:40, 50:50, 40:60, MPI and WPI at different energy densities of 11.7 J/mL,
35.1 J/mL, 58.5 J/mL, 81.9 J/mL and 117.0 J/mL. Capital and simple super-
scripts denote the significant difference between sonication times and milk
protein ratios respectively.
energy density (117.0 J/mL). Moreover, a C:W ratio of 40:60 obtained
their highest EAI at 35.1 J/mL while C:W ratio of 60:40 and WPI ob-
tained highest EAI under 58.5 J/mL and remained constant (p > 0.05)
upon increasing energy density. However, continuous increase in EAI
was observed in C:W ratio of 80:20 with increasing sono-energy while
EAI of MPI-emulsions increased up to 81.9 J/mL energy density fol-
lowed by a decrease at 117.0 J/mL.
In terms of ESI, emulsions stabilized with WPI showed the highest
ESI under all sonication times while lowest ESI observed in C:W ratio of
50:50. Interestingly, ESI of C:W ratio of 60:40, 50:50, prepared with
MPI and WPI increased gradually up to 81.9 J/mL followed by a re-
duction at 117.0 J/mL energy density.
3.5. Surface charge of emulsion
Fig. 4B shows the zeta potential values of GSO-M emulsions at dif-
ferent processing times. Emulsions stabilized with WPI showed the
highest negative zeta potential at all energy densities (p < 0.05) and
the values were not affected with an increase in the applied energy.
Similarly, the zeta potential of C:W ratio of 80:20 also remained un-
changed during sonication. In the system of C:W ratio of 40:60, in-
creases in zeta potential were observed up to 35.1 J/mL energy density
followed by constant values at higher applied energy. Moreover, the
zeta potential of C:W ratio of 60:40, 50:50 and MPI increased up to
35.1 J/mL then reduced at 117.0 J/mL applied energy.
4. Discussion
Sono-emulsification comprises a two-step mechanism. The first step
involves the eruption of dispersed phase droplets into the continuous
phase leading to smaller droplet formation which is driven by the
mechanical vibrations generated from turbulence. The second step in-
volves breaking of the droplets through the shear forces generated by
acoustic cavitation near the interface (Canselier, Delmas, Wilhelm, &
6
Food Chemistry 309 (2020) 125758
Abismail, 2002). Laplace pressure (p) which denote the energy (in the
form of shear) required by the continuous phase to deform droplets of
the dispersed phase, depends on the radius of emulsion droplets (R) and
interfacial tension (IT). Laplace pressure can be calculated using p =
(IT/2R) equation and the emulsions can be formed when the applied
shear stress is greater than the particular Laplace pressure (Canselier
et al., 2002; Walstra, 1993). Therefore, interfacial tension and the
droplet size of the emulsion provide valuable information regarding the
feasibility of emulsion formation. Moreover, the composition and the
physico-chemical characteristics in the interfacial layer are also im-
portant in determining the emulsion stability. Furthermore, zeta po-
tential relationship between EAI and ESI are key indicators for short-
term stability of emulsions.
4.1. Milk systems with C:W ratio of 80:20 and MPI
Sono-emulsification of GSO and milk protein solutions with C:W
ratio of 80:20 and MPI produced stable emulsions at energy densities of
35.1 J/mL, 58.5 J/mL, 81.9 J/mL and 117.0 J/mL. The MPI solution
and C:W ratio of 80:20 showed the equilibrium interfacial tension value
around 14.5 mN/m before the application of ultrasound (Fig. 1). DIFT
of MPI decreased with increase in sono-energy while increased in DIFT
was observed in C:W ratio of 80:20 and this may be attributed to the
sonication-induced structural changes of milk proteins. As observed in
our previous work (Silva et al., 2018), milk solutions with C:W ratio of
80:20 produced more hydrophobically mediated aggregates during
sonication which hinders the adsorption of proteins at oil-water inter-
face, resulting a higher DIFT. However, the application of ultrasound
reduce the particle size of MPI solution due to the disruption of hy-
drophobic bonds exposing more hydrophobic sites to bind at the oil-
water interface and reducing the interfacial tension easily. Reduced
interfacial tension resulted in a higher rate of oil droplet breakup
during the emulsification leading to the formation of smaller droplets
(O'Sullivan et al., 2014). As a consequence, emulsions stabilized with
MPI produced smaller droplets compared to the C:W ratio of 80:20
(Fig. 4A). An increase in applied energy density further decreased the
droplet size and create a unimodal size distribution (Fig. 3A). This may
be attributed to the higher rate of collision provided by the physical
effects generated in the liquid system such as acoustic cavitation,
acoustic streaming, mechanical vibrations, microsteaming, shear and
turbulence at higher energy densities (Abbas, Hayat, Karangwa,
Bashari, & Zhang, 2013; Ashokkumar et al., 2010).
Correspondingly, the percentage of adsorbed proteins and the sur-
face load increased as the sonication time increases resulting in more
stable emulsions at higher applied energy densities (Table 1). More-
over, the adsorbed protein layer of these two systems were primarily
consisted with the caseins indicating 69% and 62% for MPI and C:W
ratio of 80:20 respectively at 117.0 J/mL applied energy density
(Fig. 2A). Generally, a monolayer surface coverage can be observed in
the interfacial layer of caseins stabilized emulsions and 1–2 mg/m2 of
surface coverage of caseins is sufficient to produce fine stable emulsions
(Singh, 2005). Moreover, the minimum casein surface coverage was
found to be 1 mg/m2 with a 5 nm of interfacial layer thickness and the
maximum coverage reported as 3 mg/m2 with 10 nm thickness (Fang &
Dalgleish, 1993). Higher surface load values (> 3 mg/m2) observed in
the present study at higher energy densities may be attributed with the
formation of the secondary protein layer around the emulsion droplet
(Hunt & Dalgleish, 1994) or the formation of casein aggregates
(Srinivasan, Singh, & Munro, 1996). The presence of 31% and 37%
whey proteins (Fig. 2A) at the adsorbed layer of MPI and 80:20 re-
spectively at 117.0 J/mL energy density denotes the formation of ag-
gregates with un-adsorbed whey protein rather than the formation of a
secondary layer. The decrease in EAI and ESI values and increase in zeta
potential values in MPI emulsions at 117.0 J/mL energy density may be
attributed with the depletion flocculation phenomenon (Table 1). Un-
folding and surface denaturation of β-Lactoglobulin may lead to the
M. Silva, et al.
depletion flocculation of emulsion droplets (Dickinson, 2010) and it
was insensitive with particle size measurements as emulsion droplets
maintain their individual integrity during the depletion flocculation.
4.2. Milk systems with C:W ratio of 60:40 and 50:50
Stable emulsions were produced using milk protein solutions with
C:W ratio of 60:40 at 58.5 J/mL, 81.9 J/mL and 117.0 J/mL sono-en-
ergy density whereas a C:W ratio of 50:50 produced stable emulsions at
81.9 J/mL and 117.0 J/mL sono-energy density. However, there were
no differences of DIFT between untreated and sonicated (up to 117.0 J/
mL) milk solutions with C:W ratio of 60:40 and 50:50 (Fig. 1). This
suggested that the rate of the protein adsorbed at the oil–water inter-
face was approximately similar for the untreated and ultrasound treated
solutions (O'Sullivan et al., 2014). Upon increasing applied energy
density, an increase in the percentage of adsorbed proteins was ob-
served although interfacial properties were not improved (Table 1).
Therefore, it can be suggested that besides the interfacial lowering ef-
fect of milk proteins, mechanical and shear forces generated from so-
nication also greatly improve the emulsifying ability and emulsion
properties such as the formation of smaller droplets (Shanmugam &
Ashokkumar, 2014) which thereby leads to the higher adsorption of
proteins at the interface. Nevertheless, the interfacial layer of these two
systems primarily consisted of casein proteins (~60% at 117.0 J/mL
energy density – Fig. 2A) and a considerable fraction of whey proteins
remained in the continuous phase. Presence of more un-adsorbed pro-
teins in the aqueous phase during sono-emulsification induced the un-
folding and denaturation of whey proteins leading to the formation of
aggregates via hydrophobic and thiol-disulphide interactions. More-
over, denatured whey proteins show a higher affinity for κ-caseins than
the whey proteins resulting in the formation of bonds between dena-
tured whey proteins and κ-caseins than between the whey proteins
themselves (Donato, Guyomarc’h, Amiot, & Dalgleish, 2007). These
thiol-disulphide interchange reactions lead to strong crosslinking be-
tween denatured globular proteins. Furthermore, there was a possibility
to form hydrophobic interactions between non-adsorbed globular pro-
teins and interfacial proteins. As a result of these interactions, floccu-
lates and/or aggregates were formed leading to the instability of the
emulsion (Euston, Al-Bakkush, & Campbell, 2009; Euston, Finnigan, &
Hirst, 2000; Liang, Patel, Matia-Merino, Ye, & Golding, 2013;
McSweeney, Mulvihill, & O'Callaghan, 2004). Therefore, emulsions
prepared by C:W ratio of 50:50 at shorter sonication times showed layer
separation (Fig. 2B) and lowest EAI and ESI values (Table 1) although it
had the highest AP% at 1 min sonication. Upon sonication, higher en-
ergy densities helps in the mixing of oil and aqueous phases and to
overcome the Laplace pressure to form stable emulsions (Shanmugam &
Ashokkumar, 2014) with the higher surface load. However, increasing
applied energy density may lead to the instability of emulsion due to
the depletion flocculation as showed in reduced ESI (Table 1) and in-
creased zeta potential (Fig. 4B) at 117.0 J/mL applied energy density.
4.3. Milk systems with C:W ratio of 40:60
Emulsions containing milk protein solution with C:W ratio of 40:60
formed stable emulsions at 35.1 J/mL–117.0 J/mL sono-energy density.
The composition of interfacial layer (~59% whey proteins under
117.0 J/mL energy density – Fig. 2A) suggests that prolong sonication
induced the aggregation of whey proteins leading to the adsorption of
more whey protein aggregates at the interface. Lower amounts of
caseins compared to the whey proteins in the continuous phase may
form a thick and heterogeneous interface with more protruding fractal
whey proteins due to the denaturation and aggregation of whey pro-
teins. It reduces the chance of aggregation between caseins and whey
proteins (Surel et al., 2014). Presence of more aggregated structures at
the droplet interface led to the highest particle size at 11.7 J/mL energy
density compared to the other ratios (Fig. 4A). Moreover, aggregated
7
Food Chemistry 309 (2020) 125758
whey proteins in the adsorbed layer hinder the further adsorption of
proteins upon sonication due to the low availability of thio-disulphide
interchanges or hydrophobic sites for further aggregation resulting
significantly smaller AP% at 117.0 J/mL energy density compared to
the other emulsions (Table 1). This behaviour was further justified by
the EAI and zeta potential values as they remained unchanged after the
35.1 J/mL applied energy density (Table 1 & Fig. 4B).
4.4. Milk systems with WPI
Emulsions prepared by WPI alone produced stable emulsions at very
low energy density (11.7 J/mL) showing higher emulsifying ability.
Mechanical and shear properties generated through ultrasound and
reduction of interfacial tension upon sonication enhance the emulsi-
fying ability of WPI system. However, equilibration time was long for
WPI compared to the other casein containing milk protein solutions
(Fig. 1). Globular proteins such as whey proteins take a long time to
unfold and subsequently reduce the interfacial tension due to the nature
of tertiary and secondary structures compared to the flexible random-
coiled structures of the caseins (Wilde, 2000). Generally, two distinct
phases can be observed in the DIFT curves of globular proteins before
reaching to the equilibrium; first phase – a sharp decrease of DIFT due
to the diffusion and adsorption of proteins to the surface, second phase
– slow decrease in DIFT due to the conformational arrangement of
adsorbed protein at the interface. Flexible proteins such as caseins are
more effective in reducing interfacial tension due to the presence of
more non-polar groups than rigid proteins or proteins with fewer non-
polar groups (Nakai & Li-Chan, 1988). Application of ultrasound in-
duces the dissociation, denaturation and unfolding of β-Lactoglobulin,
which thereby leads to aggregation themselves (Silva et al., 2018).
Therefore, at the moment of emulsion formation, WPI solution con-
tained native, unfolded and aggregated whey proteins. As native whey
proteins are better emulsifiers than the aggregated proteins (Surel et al.,
2014), native whey proteins may preferentially adsorb to the interface
as their dimer structures (Mackie, Mingins, & Dann, 1991). Therefore,
WPI emulsions showed lower AP% (3.8%) (Table 1) and the smallest
globule size at 11.7 J/mL energy density (Fig. 4A). Upon sonication
adsorbed proteins may also undergo confirmation reorganization al-
lowing further exposure of hydrophobic sites (Zhai et al., 2010). At the
same time, prolong sonication may lead to the unfolding of unabsorbed
whey proteins which thereby improves the surface activity of un-ad-
sorbed β-lactoglobulin (Das & Kinsella, 1990) leading to the formation
of aggregation via thiol-disulphide exchanges between adsorbed and
un-adsorbed whey proteins. Therefore, a higher percentage of whey
protein adsorption could be observed as the applied energy density
increases resulting in 26.2% of AP% at 117.0 J/mL (Fig. 2A). Corre-
spondingly, emulsions stabilized with WPI showed the highest negative
zeta potential at all the sonication times and did not changed upon
ultrasonic treatment (Fig. 4B). Higher magnitude in charged showed
the higher stability of an emulsion due to the enhanced electrostatic
repulsions between droplets (Guzey & McClements, 2007). Similarly,
highest EAI and ESI observed even at 11.7 J/mL energy density com-
pared to the other emulsions (Table 1).
Sono-emulsification properties of GSO primary emulsions were in-
fluenced by the composition of the milk protein present in the aqueous
phase and the applied energy density. As shown in Table 2, EAI was
negatively correlated with particle size and positively correlated with
the percentage of adsorbed proteins. Similar to the EAI, ESI values were
also found to be negatively correlated with particle size and positively
correlated with the percentage of adsorbed proteins. Furthermore, there
was a positive correlation between EAI and ESI values. For instance,
increasing applied energy density during the production of WPI emul-
sions from 11.7 J/mL to 117.0 J/mL decreased the particle size from
1.580 µm to 0.637 µm along with the increasing EAI values ranging
from 30.68 m2/g to 34.59 m2/g. Moreover, more than a threefold in-
crease in ESI was observed at 81.9 J/mL sonication compared to the
M. Silva, et al.
Table 2
Pearson correlation coefficients (r) for emulsifying properties of GSO-M emulsions (EAI: Emulsion activity index, ESI: Emulsion stability index, PS: particle size, AP%:
adsorbed protein %).
Correlation variables Casein:whey protein ratio
80:20 60:40
EAI & PS −0.893* −0.969**
ESI & PS −0.801* −0.832*
EAI & AP 0.720* 0.726*
ESI & AP 0.917* 0.693*
EAI & ESI 0.918* 0.772*
PS & AP −0.720* −0.811*
* Correlation is significant at the 0.05 level.
** Correlation is significant at the 0.01 level.
11.7 J/mL. These relationships among emulsion properties can provide
valuable insight into the prediction of emulsion stability in future stu-
dies.
5. Conclusion
The application of low-frequency ultrasound at 20 kHz produced
stable emulsions of 10% (w/w) grape seed oil and 3.3% (w/w) milk
protein solutions. The composition of the milk proteins and the applied
energy density influence on the interfacial and emulsification proper-
ties of milk protein stabilized emulsions during the sono-emulsification.
The increase in energy density resulted in smaller droplet size with
narrow size distribution, increasing percentage of adsorbed proteins
with higher surface load, increasing emulsion activity index and
emulsion stability index. If the milk protein solution contained equal or
more caseins than the whey proteins, the interface was mainly con-
stituted by the caseins and casein-whey protein aggregates due to the
ultrasound-induced denaturation of whey proteins during the sonica-
tion leading to the aggregation via hydrophobic and thiol-disulphide
interactions. However, when the concentration of casein was less than
the whey protein, the interface is primarily covered by the whey-whey
and casein-whey aggregates. Moreover, adsorption of aggregated pro-
tein structures led to the larger droplet size. If the aqueous phase
consisted with only whey proteins, the interface was mainly covered by
the native whey proteins and whey-whey aggregates resulting smaller
particle size. Therefore, it should finally be stated that under different
energy densities, the amount of adsorbed milk proteins at the surface
could be mainly attributed to the different molecular structure and
bonding nature of caseins and/or whey proteins in the continues phase.
This may provide valuable insight to future researches in designing
milk protein-stabilized emulsions with varying milk composition
Moreover, sonication can be used as an effective emulsification method
with the careful selection of process parameters. Pilot scale experiments
are recommended under similar energy densities and compositions as
possible in order to understand the gaps between the laboratory and
scaling-up applications.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgement
The authors gratefully acknowledge the financial and technical
support from Royal Melbourne Institute of Technology, Australia and
international postgraduate research scholarship (RMIT IPRS).
8
Food Chemistry 309 (2020) 125758
50:50 40:60 MPI WPI
−0.888* −0.978** −0.945* −0.964**
−0.784* −0.697* −0.741* −0.893*
0.890* 0.690* 0.820* 0.839*
0.742* 0.949* 0.822* 0.862*
0.686* 0.559* 0.846* 0.886*
−0.753* −0.810* −0.907* −0.909*
References
Abbas, S., Hayat, K., Karangwa, E., Bashari, M., & Zhang, X. (2013). An overview of
ultrasound-assisted food-grade nanoemulsions. Food Engineering Reviews, 5(3),
139–157.
Ashokkumar, M., Bhaskaracharya, R., Kentish, S., Lee, J., Palmer, M., & Zisu, B. (2010).
The ultrasonic processing of dairy products—An overview. Dairy Science &
Technology, 90(2–3), 147–168.
Canselier, J., Delmas, H., Wilhelm, A., & Abismail, B. (2002). Ultrasound
emulsification—An overview. Journal of Dispersion Science and Technology, 23(1–3),
333–349.
Chandrapala, J., Martin, G., Zisu, B., Kentish, S., & Ashokkumar, M. (2012). The effect of
ultrasound on casein micelle integrity. Journal of Dairy Science, 95(12), 6882–6890.
Chandrapala, J., Martin, G. J. O., Kentish, S. E., & Ashokkumar, M. (2014). Dissolution
and reconstitution of casein micelle containing dairy powders by high shear using
ultrasonic and physical methods. Ultrasonics Sonochemistry, 21(5), 1658–1665.
Chandrapala, J., Zisu, B., Palmer, M., Kentish, S., & Ashokkumar, M. (2011). Effects of
ultrasound on the thermal and structural characteristics of proteins in reconstituted
whey protein concentrate. Ultrasonics Sonochemistry, 18(5), 951–957.
Dalgleish, D. (1995). Structures and properties of adsorbed layers in emulsions containing
milk proteins. In Food macromolecules and colloids (pp. 23–33).
Das, K., & Kinsella, J. (1990). Effect of heat denaturation on the adsorption of β-lacto-
globulin at the oil/water interface and on coalescence stability of emulsions. Journal
of Colloid and Interface Science, 139(2), 551–560.
Dickinson, E. (2001). Milk protein interfacial layers and the relationship to emulsion
stability and rheology. Colloids and Surfaces B: Biointerfaces, 20(3), 197–210.
Dickinson, E. (2010). Flocculation of protein-stabilized oil-in-water emulsions. Colloids
and Surfaces B: Biointerfaces, 81(1), 130–140.
Dickinson, E., & Golding, M. (1998). Influence of calcium ions on creaming and rheology
of emulsions containing sodium caseinate. Colloids and Surfaces A: Physicochemical
and Engineering Aspects, 144(1), 167–177.
Dickinson, E., & McClements, D. J. (1995). Advances in food colloids. Springer Science &
Business Media.
Donato, L., Guyomarc’h, F., Amiot, S., & Dalgleish, D. G. (2007). Formation of whey
protein/κ-casein complexes in heated milk: Preferential reaction of whey protein
with κ-casein in the casein micelles. International Dairy Journal, 17(10), 1161–1167.
Dopierala, K., Javadi, A., Krägel, J., Schano, K.-H., Kalogianni, E., Leser, M., & Miller, R.
(2011). Dynamic interfacial tensions of dietary oils. Colloids and Surfaces A:
Physicochemical and Engineering Aspects, 382(1–3), 261–265.
Euston, S., Finnigan, S., & Hirst, R. (2000). Aggregation kinetics of heated whey protein-
stabilized emulsions. Food Hydrocolloids, 14(2), 155–161.
Euston, S. R., Al-Bakkush, A.-A., & Campbell, L. (2009). Comparing the heat stability of
soya protein and milk whey protein emulsions. Food Hydrocolloids, 23(8), 2485–2492.
Fang, Y., & Dalgleish, D. G. (1993). Dimensions of the adsorbed layers in oil-in-water
emulsions stabilized by caseins. Journal of Colloid and Interface Science, 156(2),
329–334.
Fernandes, L., Casal, S., Cruz, R., Pereira, J. A., & Ramalhosa, E. (2013). Seed oils of ten
traditional Portuguese grape varieties with interesting chemical and antioxidant
properties. Food Research International, 50(1), 161–166.
Guzey, D., & McClements, D. J. (2007). Impact of electrostatic interactions on formation
and stability of emulsions containing oil droplets coated by β-lactoglobulin− pectin
complexes. Journal of Agricultural and Food Chemistry, 55(2), 475–485.
Hunt, J. A., & Dalgleish, D. G. (1994). Adsorption behaviour of whey protein isolate and
caseinate in soya oil-in-water emulsions. Food Hydrocolloids, 8(2), 175–187.
Liang, Y., Patel, H., Matia-Merino, L., Ye, A., & Golding, M. (2013). Structure and stability
of heat-treated concentrated dairy-protein-stabilised oil-in-water emulsions: A sta-
bility map characterisation approach. Food Hydrocolloids, 33(2), 297–308.
Mackie, A., Mingins, J., & Dann, R. (1991). Preliminary studies of β-lactoglobulin ad-
sorbed on polystyrene latex. Food Polymers, Gels and Colloids (pp. 96–112). Elsevier.
Madawala, S., Kochhar, S., & Dutta, P. (2012). Lipid components and oxidative status of
selected specialty oils. Grasas y Aceites, 63(2), 143–151.
McClements, D. J. (2015a). Encapsulation, protection, and release of hydrophilic active
components: Potential and limitations of colloidal delivery systems. Advances in
Colloid and Interface Science, 219, 27–53.
McClements, D. J. (2015b). Food emulsions: Principles, practices, and techniques. CRC Press.
M. Silva, et al.
McSweeney, S. L., Mulvihill, D. M., & O'Callaghan, D. M. (2004). The influence of pH on
the heat-induced aggregation of model milk protein ingredient systems and model
infant formula emulsions stabilized by milk protein ingredients. Food Hydrocolloids,
18(1), 109–125.
Nakai, S., & Li-Chan, E. (1988). Hydrophobic interactions in food systems. CRC Press.
O'Sullivan, J., Arellano, M., Pichot, R., & Norton, I. (2014). The effect of ultrasound
treatment on the structural, physical and emulsifying properties of dairy proteins.
Food Hydrocolloids, 42, 386–396.
Parkinson, E. L., & Dickinson, E. (2004). Inhibition of heat-induced aggregation of a beta-
lactoglobulin-stabilized emulsion by very small additions of casein. Colloids Surf B
Biointerfaces, 39(1–2), 23–30.
Pearce, K. N., & Kinsella, J. E. (1978). Emulsifying properties of proteins: Evaluation of a
turbidimetric technique. Journal of Agricultural and Food Chemistry, 26(3), 716–723.
Pichot, R. (2012). Stability and characterisation of emulsions in the presence of colloidal
particles and surfactants. University of Birmingham.
Ranadheera, C. S., Liyanaarachchi, W. S., Chandrapala, J., Dissanayake, M., & Vasiljevic,
T. (2016). Utilizing unique properties of caseins and the casein micelle for delivery of
sensitive food ingredients and bioactives. Trends in Food Science & Technology, 57,
178–187.
Shanmugam, A., & Ashokkumar, M. (2014). Ultrasonic preparation of stable flax seed oil
emulsions in dairy systems – Physicochemical characterization. Food Hydrocolloids,
39, 151–162.
Shanmugam, A., Chandrapala, J., & Ashokkumar, M. (2012). The effect of ultrasound on
the physical and functional properties of skim milk. Innovative Food Science &
Emerging Technologies, 16, 251–258.
Food Chemistry 309 (2020) 125758
Silva, M., Zisu, B., & Chandrapala, J. (2018). Influence of low-frequency ultrasound on
the physico-chemical and structural characteristics of milk systems with varying
casein to whey protein ratios. Ultrasonics Sonochemistry, 49, 268–276.
Singh, H. (2005). Milk protein functionality in food colloids. Food Colloids-Interactions,
Microstructure and Processing179–193.
Srinivasan, M., Singh, H., & Munro, P. A. (1996). Sodium caseinate-stabilized emulsions:
Factors affecting coverage and composition of surface proteins. Journal of Agricultural
and Food Chemistry, 44(12), 3807–3811.
Surel, C., Foucquier, J., Perrot, N., Mackie, A., Garnier, C., Riaublanc, A., & Anton, M.
(2014). Composition and structure of interface impacts texture of O/W emulsions.
Food Hydrocolloids, 34, 3–9.
Walstra, P. (1993). Principles of emulsion formation. Chemical Engineering Science, 48(2),
333–349.
Wilde, P. (2000). Interfaces: Their role in foam and emulsion behaviour. Current Opinion
in Colloid & Interface Science, 5(3–4), 176–181.
Ye, A. (2008). Interfacial composition and stability of emulsions made with mixtures of
commercial sodium caseinate and whey protein concentrate. Food Chemistry, s110(4),
946–952.
Ye, A., & Singh, H. (2000). Influence of calcium chloride addition on the properties of
emulsions stabilized by whey protein concentrate. Food Hydrocolloids, 14(4),
337–346.
Zhai, J., Miles, A. J., Pattenden, L. K., Lee, T.-H., Augustin, M. A., Wallace, B. A., ...
Wooster, T. J. (2010). Changes in β-lactoglobulin conformation at the oil/water in-
terface of emulsions studied by synchrotron radiation circular dichroism spectro-
scopy. Biomacromolecules, 11(8), 2136–2142.