Você está na página 1de 11

Journal of Food Engineering 164 (2015) 10–20

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Physical and oxidative stability of fish oil nanoemulsions produced

by spontaneous emulsification: Effect of surfactant concentration
and particle size
Rebecca M. Walker a, Eric A. Decker a,b, David J. McClements a,b,⇑
Department of Food Science, University of Massachusetts, Amherst, MA 01003, USA
Department of Biochemistry, Faculty of Science, King Abdulaziz University, P.O. Box 80203, Jeddah 21589, Saudi Arabia

a r t i c l e i n f o a b s t r a c t

Article history: Nanoemulsion-based delivery systems offer many potential benefits for incorporating omega-3 oils into
Received 9 February 2015 foods and beverages. Nanoemulsions are gaining popularity because of their ease of preparation, small
Received in revised form 19 April 2015 particle size, relatively high stability, and production of optically transparent emulsions. In this study,
Accepted 28 April 2015
spontaneous emulsification, a low-energy method, was used to fabricate fish oil nanoemulsions. The
Available online 29 April 2015
influence of surfactant-to-oil-ratio on particle size and physical stability was evaluated. Optically trans-
parent nanoemulsions were formed and maintained physical stability at 37 °C for 14 days. Furthermore,
the effect of particle size and surfactant concentration on oxidative stability of these nanoemulsions was
Fish oil
Spontaneous emulsification
compared to emulsions produced by microfluidizer, a high-energy method. These nanoemulsions had
Nanoemulsion similar oxidative stabilities at 55 °C for 14 days. These results demonstrate that spontaneous emulsifica-
Lipid oxidation tion can produce fish oil nanoemulsions that are physically stable and oxidize at similar rates as tradi-
Lemon oil tionally prepared nanoemulsions, and therefore may be suitable for fortifying clear food systems.
Microfluidizer Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction delivery systems are particularly suitable for incorporating fish oils
into functional food products.
Fish oil is an excellent source of long chain polyunsaturated Nanoemulsions, a class of emulsion-based delivery systems,
fatty acids (PUFAs), such as eicosapentaenoic acid (EPA, 20:5 n-3) have been of particular interest lately because of their simple fab-
and docosahexaenoic acid (DHA, 22:6 n-3) (Maki et al., 2014). rication, high physical stability, and high bioavailability
Consumption of adequate levels of fish oil has been shown to pro- (McClements, 2011). By definition, nanoemulsions have a mean
vide health benefits associated with brain development, inflamma- droplet radii below 100 nm, and may become optically transparent
tion, and cardiovascular disease (Kris-Etherton et al., 2009). In the at sufficiently small particle sizes (Mason et al., 2006; Tadros et al.,
United States, the 2010 Dietary Guidelines for Americans (USDA) 2004). In contrast to microemulsions, which are thermodynami-
recommends the consumption of 250 mg of EPA and DHA per cally stable systems, nanoemulsions are thermodynamically unsta-
day through the means of 8 oz (227 g) of a variety of seafood a ble systems but can be designed to be kinetically stable (Anton
week. Unfortunately, Americans are falling short of this recom- et al., 2008; Anton and Vandamme, 2011; McClements, 2012).
mendation with a current consumption of 3.5 oz (99 g) of seafood Nanoemulsions can be fabricated by high or low-energy methods.
per week, mostly from sources low in omega-3 FAs (Kris-Etherton High-energy methods use specialized mechanical devices to break-
et al., 2009). This under-consumption of seafood may be attributed down the droplets into very fine particles, such as microfluidizers
to taste, price, contamination concerns (such as heavy metals), and (MF), high pressure valve homogenizers, or sonicators
availability (Glanz et al., 1998; Kennedy et al., 2012; Racine and (McClements, 2011; McClements and Rao, 2011; Tadros et al.,
Deckelbaum, 2007). As a result, there is a need to develop alterna- 2004). In contrast, low-energy methods are able to spontaneously
tive sources for omega-3 FAs in consumer’s diets. Emulsion-based form very fine droplets as a result of controlled changes in the
environment or solution conditions (McClements and Rao, 2011;
Solans and Sole, 2012). The interest in low-energy methods for cer-
tain applications is increasing because of their lower manufactur-
⇑ Corresponding author at: Department of Food Science, University of Mas-
ing costs, simple production methods, and ability to create
sachusetts, Amherst, MA 01103, USA.
E-mail address: mcclements@foodsci.umass.edu (D.J. McClements). smaller particle sizes than high-energy methods (McClements

0260-8774/Ó 2015 Elsevier Ltd. All rights reserved.
R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20 11

and Rao, 2011). Spontaneous emulsification (SE) is one of the sim- solvents and reagents were of analytical grade or higher. Double
plest low-energy methods to implement since it only involves the distilled water was used to prepare all solutions.
addition of one phase into another phase with continuous stirring
to spontaneously form a nanoemulsion (Anton and Vandamme, 2.2. Emulsion preparation
2009; Saberi et al., 2013). Typically in SE, the organic phase con-
sisting of oil and surfactant is added to the aqueous phase with 2.2.1. Low-energy method: Spontaneous emulsification
mild stirring. Nanoemulsions were prepared by spontaneous emulsification
This study will focus on the potential of spontaneous emulsifi- (SE) for physical and oxidation stability evaluation. The organic
cation to fabricate fish oil nanoemulsions that are suitable for phase consisted of a mixture of fish oil (5 wt%) and lemon oil
application in clear beverages. As part of this study, the nanoemul- (5 wt%), which were stirred at 750 rpm for 15 min, and then stirred
sions created using this low-energy method will be compared to for an additional 30 min after adding non-ionic surfactant (2.5–
those produced using a high-energy method (MF) to highlight 20 wt% Tween 80). The aqueous phase was buffer (70–87.5 wt%)
the advantages and limitations of these different approaches. consisting of 0.8 wt% citric acid and 0.08 wt% sodium benzoate at
Previous studies haves shown that microfluidization is a particu- pH 3.0, in order to simulate the aqueous phase of a beverage sys-
larly efficient high-energy method of producing nanoemulsions tem. In this method, the organic phase was added to an aqueous
containing small lipid droplets (Jafari et al., 2008, 2007, 2006). phase using an automatic pipette (Ranin 10 mL E4 XLS, Mettler-
Both the physical and chemical stability of nanoemulsions are Toledo International Inc., Columbus, OH, USA) while stirring at
obstacles that must be addressed when producing foods fortified 500 rpm for 15 min.
with omega-3 fatty acids (McClements et al., 2007), and so these
issues will be evaluated in this research. 2.2.2. High-energy method: Microfluidizer
Nanoemulsions were also prepared by microfluidization (MF)
2. Materials and methods for oxidation stability evaluation. Fish oil (10 wt%) and lemon oil
(10 wt%) were mixed for 15 min at 750 rpm. Buffer (78 wt%) was
2.1. Materials mixed with Tween 80 (2 wt%) for 30 min at 750 rpm. The two
phases were added together and mixed for 2 min with a hand
Fish oil (FO) (Ropufa 30 n-3 food oil) was provided by DSM mixer (Bamix ESGE Ltd., Switzerland) to form a course emulsion.
Nutritional Products Ltd. (Basal, Switzerland). The oil was com- Samples were passed through a microfluidizer (M-110L,
posed of 101 mg of EPA/g of oil, 148 mg of DHA/g oil, and Microfluidics, Newton, MA) 3 times at 12,000 PSI.
312 mg of total n-3 PUFA/g of oil. Lemon oil (LO) was kindly
donated by Citrus & Allied Essences (Lake Success, NY, USA). The 2.2.3. Post-production alterations of emulsions
supplier reported the chemical composition as determined by gas All emulsions were diluted to 1 wt% oil (0.5 wt% FO and 0.5 wt%
chromatography (Table 1). Non-ionic surfactant, polysorbate 80 LO) with buffer solution and then stirred for 5 min at 300 rpm.
(Tween 80), sodium benzoate, thiobarbituric acid (TBA), butylated Finished emulsions were held in 50 mL disposable centrifuge
hydroxytoluene (BHT), 1,1,3,3-tetraethoxypropane (TEP), barium polypropylene tubes (Fisher Scientific, Pittsburg, PA, USA). Each
chloride, iron (II) sulfate heptahydrate, hydrochloric acid, and formulation was made in duplicate.
cumene hydroperoxide were purchased from Sigma–Aldrich Co. For the emulsions used in oxidation studies, additional surfac-
(St. Lois, MO, USA). Citric acid, isooctane, 1,2-propanol, methanol, tant was added during the dilution stage to evaluate the effect of
and butanol were purchased from Fisher Scientific (Waltham, surfactant and particle size on oxidation. For these emulsions,
MA, USA). Trichloroacetic acid (TCA) and ammonium thiocyanate Tween 80 was mixed with the volume of buffer used for dilution
were purchased from Acros Organics (Geel, Belgium). Ethanol at 750 rpm for 30 min. This solution was added to the emulsion
was purchased from Pharmco-AAPER (Brookfield, CT, USA). All to dilute to 1 wt% oil (0.5 wt% FO and 0.5 wt% LO) and stirred at
300 rpm for 5 min. Iron (100 lM) as FeSO4 was also added to all
emulsions used in the oxidation studies to accelerate the lipid oxi-
Table 1 dation reaction. Emulsions were observed using optical micro-
Concentration of constituents in threefold (3) lemon oil, provided by Citrus & Allied
scopy on day 0 and 14 of the oxidation experiments (Nikon
Essences (Lake Success, NY).
Eclipse 80i, Nikon Instrument Inc., Melville, NY). Each formulation
Constituent Concentration in lemon oil (%) was made in duplicate.
a-Thujene 0.00
a-Pinene 0.70 2.3. Surfactant concentration
Camphene 0.00
Sabinene 0.30
b-Pinene 4.90
The effect of surfactant concentration in the SE nanoemulsions
Myrcene 0.80 was evaluated by varying the surfactant-to-oil ratio (SOR) while
Octanal 0.01 keeping the total amount of oil (fish oil and lemon oil) constant
Limonene 63.00 (10 wt%).
a-Phellandrene 0.02
c-Terpinene 14.00 %SOR ¼ 100  ms =mo
Terpinolene 1.80
Linalool 0.60 where ms is the mass of the surfactant and mo is the total mass of
Citronellal 0.30 the oil phase.
a-Terpineol 0.30
Neral (citral B) 2.38
Geranial (citral A) 3.84
2.4. Particle size measurements
Neryl acetate 0.97
Geranyl acetate 0.60 The particle size distribution (PSD) of all emulsions was mea-
(E)-caryophyllene 0.90 sured by either dynamic (Zetasizer Nano ZS, Malvern
Trans-a-bergamotene 0.90
Instruments, Malvern, UK) or static light scattering instruments
b-Bisabolene 1.40
(Mastersizer 2000, Malvern Instruments, Malvern, UK). Static light
Total (%) 97.72
scattering was used to measure the size of the droplets in
12 R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20

emulsions containing relatively large droplets (d > 1 lm), and in 4. Results and discussion
this case the particle size was reported as d32. On the other hand,
dynamic light scattering was used to analyze emulsions containing For the sake of convenience, the term ‘‘emulsions’’ is used to
relatively small droplets (d < 1 lm), and the particle size was refer to both nanoemulsions (d < 200 nm) and conventional emul-
reported as the Z-average diameter and polydispersity index sions (d > 200 nm) in the remainder of this study, since both types
(PDI). PDI is a measure of the narrowness of the size distribution; of systems could be formed depending on the initial composition
values 60.1 designate a very narrow size distribution (Saberi of the system. The fish oil (FO) was blended with lemon oil (LO)
et al., 2013). All samples were measured in duplicate. prior to homogenization since previous studies have shown that
this flavor oil has a number of benefits in forming nanoemulsions:
(i) it facilitates emulsion formation (by reducing oil phase viscosity
2.5. Turbidity measurement and interfacial tension); (ii) it can mask any off-flavors; (iii) and it
may contain natural antioxidants (Gulotta et al., 2014a; Rao and
The turbidity of samples was determined using a Thermo McClements, 2011, 2013).
Scientific Evolution Array UV–Visible Spectrophotometer
(Waltham, MA, USA). The turbidity was measured at 600 nm at
4.1. Effect of surfactant concentration on physical stability
the temperature at which the samples were being held. All samples
were measured in duplicate.
Emulsions were prepared using spontaneous emulsification
with 10 wt% total oil phase (50 wt% FO and 50 wt% LO) and differ-
ent surfactant-to-oil ratios (SOR). These emulsions were then
2.6. Oxidation measurements diluted to 1 wt% total oil phase and held at 5, 20, and 37 °C for
14 days to evaluate their physical stability. Surfactant concentra-
2.6.1. Peroxide value tion had a large impact on the mean particle diameter of the emul-
Emulsions were held in 50 mL disposable centrifuge polypropy- sions: as the SOR increased, the mean particle diameter decreased
lene tubes (Fisher Scientific, Pittsburg, PA, USA) and incubated in
the dark at 55 °C for 14 days, with measurements being taken
every 2 days. Lipid hydroperoxides were measured using a method A 10000
adapted from Shantha and Decker (1994). Lipids from the emulsion Day 0
9000 Day 1
were extracted by adding the emulsion (0.3 mL) to a 3:1 v/v mix-
Mean particle diameter (nm)

Day 7
ture of isooctane/2-propanol (1.5 mL) and vortexing the mixture 8000
Day 14
for 10 s, 3 times followed by centrifugation (1000g) for 2 min. 7000
The top layer (0.2 mL) was mixed with 2:1 v/v methanol/1-butanol
(2.8 mL), followed by 30 lL of 1:1 v/v 3.94 M ammonium thio- 6000
cyanate/ferrous iron solution (solution prepared by mixing equal 5000
amounts of 0.132 M BaCl2 and 0.144 M FeSO4). After 20 min, absor-
bance was measured at 510 nm using an UV–visible spectropho-
tometer (Ultraspec 3000 pro, Biochrom Ltd., Cambridge, UK). 3000
Hydroperoxides were calculated as mM cumene hydroperoxide 2000
using a cumene hydroperoxide standard curve at concentrations
of 0–0.5 mM. All samples were measured in triplicate.
0.00 0.50 1.00 1.50 2.00
2.6.2. Thiobarbituric acid-reactive substances (TBARS)
TBARS samples were held in 50 mL disposable centrifuge
polypropylene tubes (Fisher Scientific, Pittsburg, PA, USA) at B 9000
5 °C
55 °C in the dark and were measured every 2 days for 14 days 8000 20 °C
using the method of McDonald and Hultin (1987). Emulsion
Mean particle diameter (nm)

37 °C
(1.0 mL) and TBA reagent (2.0 mL) (15% w/v trichloroacetic acid, 7000
0.375% w/v thiobarbituric acid, and 0.25 M HCl with 2% BHT in
ethanol solution) were vortexed in glass test tubes with screw
caps. The tubes were placed in a water bath (90 °C) for 15 min, 5000
and then moved to a room temperature water bath to cool for
10 min. The tubes were centrifuged (1000g) for 15 min, then sat
for 10 more minutes. The absorbance of the supernatant was mea- 3000
sured at 532 nm. Concentrations of TBARS were calculated as nM
using a standard curve of TEP at concentrations between 0 and
20 nM. All samples were measured in triplicate. 1000

0.00 0.50 1.00 1.50 2.00
3. Experimental design and data analysis SOR

All measurements were done in duplicate or triplicate and Fig. 1. Influence of surfactant-to-oil ratio (SOR) on the mean particle diameter of
results are reported as mean ± standard deviation. Statistical anal- oil-in-water emulsions produced by spontaneous emulsification (SE) (A) after being
held at 20 °C for 14 days and (B) on day 3 at all temperatures. All systems were
ysis was carried out by analysis of variance (ANOVA) and Tukey made with a total oil phase content of 10 wt% (50% fish oil (FO) and 50% lemon oil
test with confidence interval of 95% using Minitab 16 software (LO)) and were diluted to 1 wt% total oil phase before measurement. Particle sizes
(Minitab Inc., State College, PA, USA). on day 0 were measured immediately after producing the emulsion.
R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20 13

(Fig. 1). The dependence of the mean particle size on SOR did not cost concerns, they may be suitable for application in the pharma-
depend strongly on storage time (Fig. 1a) or holding temperature cological industry, especially for topical treatments.
(Fig. 1b). Previous studies have also reported similar trends using The turbidity of the emulsions was monitored throughout their
low-energy methods to prepare emulsions, such as spontaneous storage to evaluate changes in their opacity. The turbidity of the
emulsification and emulsion phase inversion (Gulotta et al., nanoemulsions decreased as the SOR increased for all holding tem-
2014b; Mayer et al., 2013; Ostertag et al., 2012; Saberi et al., peratures and storage times (Fig. 3). This was expected, as the
2013; Yang et al., 2012). emulsions should become less turbid and more transparent as
After fabrication of the emulsions, the mean particle diameters the mean particle size decreased. Emulsions appear opaque when
ranged from 106 to 7800 nm for SORs ranging from 1.75 to 0.5, their particle size is comparable to the wavelength of light and
respectively (Fig. 1). Emulsions with SORs of 0.63 and below were the light is scattered strongly (McClements, 2002). In contrast, they
significantly larger than those with SORs of 0.75 and above appear only slightly turbid or transparent when their particle sizes
(p < 0.05). However, true nanoemulsions (d < 200 nm) were only are appreciably smaller than the wavelength of light (Mason et al.,
formed with SORs from 1.00 to 1.75 (Mason et al., 2006; Tadros 2006).
et al., 2004). These particle size results agree with those of a recent The overall dependence of turbidity on SOR for emulsions pre-
study that also prepared fish oil nanoemulsions using spontaneous pared using the spontaneous emulsification method did not
emulsification (Gulotta et al., 2014b). The fact that there was no depend strongly on storage time (Fig. 3a) or holding temperature
significant change in the mean particle size of the emulsions after (Fig. 3b). Emulsions with SORs of 0.75 and below were significantly
they were held at 5, 20, and 37 °C for 14 days, suggests that they more turbid than those with SORs of 1.0 and above (p < 0.05). The
were stable against droplet growth from flocculation, coalescence, origin of this effect can be attributed to the observed differences in
and Ostwald ripening (McClements and Rao, 2011). droplet size (Fig. 1). The droplets formed at lower SORs (0.25–0.75)
Only the particle size distributions of the emulsions held at are much larger than those formed at higher SORs (1.00–1.75), and
20 °C for 14 days are shown since these were representative of therefore they scatter light more strongly leading to a higher tur-
the other holding temperatures and storage times (Fig. 2). At bidity. The turbidity of many of the emulsions was lower after
SORs from 1.25 to 1.75, emulsions containing small droplets with 14 days storage than immediately after preparation (Fig. 3a),
a narrow particle size distribution were formed. At higher SORs
the emulsions still had a monomodal particle size distribution,
but the majority of droplets were much larger (>10 lm). It is
important to note that during the fabrication of emulsions using
A 5.0
4.5 SOR 1.75
high SORs (P1.75) gel-like globules were observed by eye and
SOR 1.50
under an optical microscope (data not shown). At SOR values of 4.0 SOR 1.25
2.00 and above, these gel-like structures did not dissolve during SOR 1.00
Absorbance (cm-1)

the 15 min period after preparing the emulsions, and so these sam- SOR 0.75
ples were not analyzed further. Some gel-like structures were also 3.0 SOR 0.63
formed at SOR 1.75, but fewer were observed and they tended to SOR 0.50
SOR 0.25
disappear after the emulsions were diluted to 1 wt% oil. We
hypothesize that these were liquid–crystalline regions formed 2.0
from surfactant, oil, and water at relatively high surfactant levels. 1.5
Further research is needed to evaluate the precise range of condi-
tions where these gel-like structures form, and to develop effective
strategies to avoid or disrupt them, if this is possible. While 0.5
emulsions with such high surfactant concentrations may not be 0.0
of commercial interest to the food industry because of legal and 0 5 10 15

B 5.0
5 °C
160 20 °C
1.75 4.0 37 °C
Absorbance (cm-1)

1.50 3.5
Relative Volume (%)

120 3.0
100 2.5
80 2.0
0.75 1.5
0.63 1.0
20 0.50
0 0.25 0.25 0.75 1.25 1.75
10 100 1000 10000 100000 SOR
Particle Diameter (nm)
Fig. 3. Influence of SOR on the turbidity of oil-in-water emulsions produced by SE
Fig. 2. Influence of SOR on the particle size distribution of oil-in-water emulsions (A) after being held at held at 20 °C for 14 days and (B) on day 3 at all temperatures.
produced by SE held at 20 °C on day 14. All systems were made with a total oil All systems were made with a total oil phase content of 10 wt% (50% FO and 50% LO)
phase content of 10 wt% (50% FO and 50% LO) and were diluted to 1 wt% total oil and were diluted to 1 wt% total oil phase before measurement. Turbidity on day 0
phase before measurement. was measured immediately after producing the emulsion.
14 R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20

which suggests that there may have been a decrease in droplet SOR = 1.75 (Fig. 5a). Interestingly, the droplet growth rate
number or size. A possible explanation for this phenomenon is sol- depended on both the initial particle size and the homogenization
ubilization of some of the oil phase in surfactant micelles during approach used. Droplet growth was fastest for the low-energy
storage. emulsion to which surfactant was added after emulsion formation,
Additional information about the influence of SOR on the stor- which initially had an intermediate particle size. On the other
age stability of the emulsions was obtained from visual observa- hand, droplet growth was slowest for the high-energy emulsion,
tions of their appearance (Fig. 4). For the sake of brevity, only which initially had the largest particle size. This result suggests
images from day 0 and 14 for emulsions stored at 37 °C are shown that the initial structural organization of the oil and surfactant
since they gave similar trends as the other samples, but exhibited molecules within the emulsions played an important role in deter-
the largest overall change in appearance. Immediately after fabri- mining their stability to droplet growth.
cation (day 0), only the emulsions with SORs from 1.25 to 1.75
were optically transparent, whereas emulsions with lower surfac-
tant concentrations were opaque. After 14 days storage, the range Influence of surfactant concentration. We also examined the
of SOR levels where emulsions appeared transparent at higher sur- influence of the SOR on the stability of emulsions that initially had
factant values increased, which again suggested that there was a similar particle sizes (Fig. 5b). These emulsions were all fabricated
reduction in the number or size of the oil droplets during storage using the microfluidizer (MF), but different amounts of surfactant
(possibly due to solubilization). In addition, there was evidence were added after production. There was little change in the mean
of an opaque layer on the top of the emulsions with the lower particle diameter during storage of the MF emulsion containing
SOR levels, which can be attributed to rapid creaming of the large no surfactant added after homogenization (MF (0.10)), which sug-
fat droplets. gests that it was relatively stable to droplet growth. On the other
hand, there was an appreciable increase in the mean droplet diam-
4.2. Effect of particle size and surfactant concentration on lipid eter of the two emulsions containing additional surfactant, with
oxidation the growth rate increasing with increasing surfactant concentra-
tion. A similar result was obtained for emulsions produced by
In this series of experiments, the influence of droplet size and spontaneous emulsification (SE): the droplet diameter increased
surfactant concentration on the oxidative stability of fish oil-in- by 64% after 14 days storage for the SE emulsion with SOR 1.00
water emulsions prepared by low-energy (SE) and high-energy (SE (1.00)), but by 117% for the SE emulsion with added surfactant
(MF) methods was examined. Emulsions with different droplet (SE+S (1.75)). These results suggest that the presence of free surfac-
sizes were prepared using the low-energy method by varying the tant within the aqueous phase of the emulsions increased the dro-
SOR (Fig. 1). However, the fact that these emulsions had different plet growth rate. There are a number of potential reasons for this
surfactant levels may also have influenced their oxidative stability. phenomenon: (i) micelles formed by non-ionic surfactants are
For this reason, we added extra surfactant to some of the emul- known to generate an osmotic attraction to oil droplets due to a
sions after they were prepared so that we could produce emulsions depletion effect, which can promote coalescence (McClements,
that had different particle sizes, but similar surfactant concentra- 1994); (ii) micelles can solubilize and transport oil molecules
tions. A similar procedure was also carried out for the emulsions between emulsion droplets, which can enhance Ostwald ripening
prepared using the high-energy method. These samples could then (Binks et al., 1999).
be used to disentangle the effects of droplet size from those of sur- In summary, the stability of the oil droplets to growth during
factant concentration on lipid oxidation. A summary of the differ- storage appears to depend on at least three factors: initial particle
ent samples prepared for this study is given in Table 2. size; total surfactant concentration; and preparation method (low-
Initially, low-energy emulsions were prepared using SORs of energy versus high-energy). The frequency of droplet collisions is
1.00 (‘‘SE (1.00)’’) and 1.75 (‘‘SE (1.75)’’) because they gave rela- known to decrease with increasing droplet size, whereas the coa-
tively stable emulsions containing large and small droplet sizes, lescence efficiency is known to increase (Capek, 2004). The rate
respectively (Table 2). Additional surfactant was then added to a of Ostwald ripening is known to increase with decreasing particle
portion of the SE (1.00) to produce an emulsion that contained rel- size, increasing polydispersity, and increasing free surfactant con-
atively large droplets and a high surfactant concentration (‘‘SE+S centration (Kabalnov, 2001; Kabalnov and Shchukin, 1992).
(1.75)’’). The diameter was controlled by adding all the surfactant Consequently, the dependence of droplet growth on these factors
prior to emulsion formation, whereas the surfactant level was con- may be quite complex.
trolled by adding some of the surfactant after emulsion formation. In all the emulsions produced using the low-energy method,
For the sake of comparison, a high-energy emulsion was also pre- there was a considerable increase in the mean particle diameter
pared with an SOR of 0.10 using a microfluidizer (‘‘MF (0.10)’’). during storage (Fig. 5a). This increase was not observed in the
Additional surfactant was also added to some of this emulsion to experiments on the influence of SOR on emulsion stability reported
prepare final samples that had SOR values similar to those found in Section 4.1, which can be attributed to the lower storage tem-
in the low-energy emulsions, ‘‘MF+S (1.00)’’ and ‘‘MF+S (1.75)’’ perature and lack of iron in those studies. Elevated temperatures
(Table 2). In naming these emulsions, the ‘‘+S’’ indicates that are known to promote droplet coalescence in emulsions stabilized
samples had surfactant added to them after fabrication and the by non-ionic surfactants due to dehydration of their head groups.
numbers in parentheses indicate the final SOR of the emulsion. For example, Gulotta et al. (2014b) conducted a thermal stability
The prepared emulsions were found to be relatively stable to analysis on nanoemulsions that contained 10% oil (50% FO and
lipid oxidation when stored under ambient conditions. We 50% LO), 10% Tween 80, and 80% aqueous phase. This experiment
therefore accelerated the lipid oxidation reaction by adding a pro- showed that rapid droplet growth occurred between 50 and
oxidant (100 lM iron) and holding them at an elevated tempera- 68 °C, which was attributed to coalescence caused by surfactant
ture (55 °C) during storage. head group dehydration. Absorbance measurements and visual
observations were also made on the same samples during storage,
4.2.1. Physical stability of emulsions during oxidation which showed that there was an increase in turbidity in the emul- Influence of initial particle size. We examined the influence sions that underwent droplet growth (data not shown), as would
of the initial particle size on the stability of the emulsions to dro- be expected due to the increase in the light scattering ability of
plet growth during storage at the same surfactant level, i.e. the larger droplets.
R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20 15

SOR 0.25 SOR 0.50 SOR 0.63 SOR 0.75 SOR 1.00 SOR 1.25 SOR 1.50 SOR 1.75

SOR 0.25 SOR 0.50 SOR 0.63 SOR 0.75 SOR 1.00 SOR 1.25 SOR 1.50 SOR 1.75

SOR 0.25 SOR 0.50 SOR 0.63 SOR 0.75 SOR 1.00 SOR 1.25 SOR 1.50 SOR 1.75

SOR 0.25 SOR 0.50 SOR 0.63 SOR 0.75 SOR 1.00 SOR 1.25 SOR 1.50 SOR 1.75

Fig. 4. Influence of SOR on the appearance of oil-in-water emulsions produced by SE (A) on day 0 at 20 °C and (B) on day 14 held at 37 °C. All systems were made with a total
oil phase content of 10 wt% (50% FO and 50% LO) and were diluted to 1 wt% total oil phase before measurement. Images on day 0 were taken immediately after producing the
16 R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20

Table 2
Comparison of the composition and structure of different nanoemulsions used in the oxidation studies. All of the emulsions contained the same final level of oil (1 wt%). Key:
SE = spontaneous emulsification (low energy); MF = Microfluidization (high energy); +S = surfactant was added after fabrication. Numbers in parentheses indicate the final SOR of
the emulsion.

Sample name Preparation method Fabrication SOR Final SOR Initial droplet diameter (nm)
SE (1.00) SE – all surfactant added initially 1.00 1.00 132.7 ± 6.4
SE (1.75) SE – all surfactant added initially 1.75 1.75 96.5 ± 2.0
SE+S (1.75) SE – some surfactant added to SE 1.00 after formation 1.00 1.75 136.0 ± 3.8
MF (0.10) MF – all surfactant added initially 0.10 0.10 151.8 ± 2.1
MF+S (1.00) MF – some surfactant added to MF 0.10 after formation 0.10 1.00 161.1 ± 1.8
MF+S (1.75) MF – some surfactant added to MF 0.10 after formation 0.10 1.75 166.5 ± 1.8

storage, which may have occurred due to the generation of some

A 300 SE (1.75)
SE+S (1.75)
free fatty acids caused by hydrolysis of the triglyceride molecules
MF+S (1.75) during storage. The emulsion produced using the microfluidizer
Mean Particle Diameter (nm)

250 with no surfactant added afterwards (MF (0.10)) had the lowest
concentration of free surfactant and a slightly more negative
charge (1.5 and 2.0 mV on days 0 and 14, respectively) than
200 the other emulsions (0.9 and 1.2 mV), which may have been
because the surfactant micelles in the higher surfactant emulsions
contributed to the signal used to calculate the f-potential.
150 Alternatively, any anionic amphiphilic components arising from
the oil phase (such as free fatty acids) may have been distributed
amongst a large number of surfactant micelles in the samples con-
taining added surfactant.
Microscopic images of the emulsions on days 0 and 14 were
50 used to provide further information about their structural proper-
0 5 10 15 ties (Fig. 6). Initially, there was no evidence of any visible particles
Storage Time (Days) in any of the samples, which is to be expected given the small par-
ticle sizes of the emulsions. Nevertheless, there was evidence of
B 240 some multiple emulsions within the SE emulsion with added sur-
MF (0.10)
MF+S (1.00)
factant (SE+S (1.75)) on day 14. These results suggest that some
220 MF+S (1.75) complex, non-equilibrium structures were formed in the emul-
Mean Particle Diameter (nm)

sions when excess surfactant was added to the SE emulsion. In

addition, transparent fragments were observed in both SE emul-
sions with SOR 1.75 (SE (1.75) and SE+S (1.75)) (Fig. 6c). We
attempted to study these fragments using cross-polarized optical
180 microscopy, but they did not appear to polarize the light because
they remained dark against a black background (data not shown),
suggesting that they were not crystalline (Fanun, 2008; Yang et al.,
2012). At this time, it is unclear what these fragments may be and
further investigation is needed. Nevertheless, the microscopy
140 images clearly show that different structures are present in emul-
sions that have similar overall compositions (SOR = 1.75), which
120 highlights the importance of the preparation method used on their
0 5 10 15 structural and physicochemical properties.
Storage Time (Days)

Fig. 5. Effect of (A) initial mean particle diameter and (B) surfactant on droplet 4.2.2. Oxidative stability of emulsions
growth of emulsions made by MF and SE and held at 55 °C with added iron. All Lipid oxidation in emulsions typically occurs at the oil–water
emulsions were diluted to 1 wt% total oil phase. Key: SE = spontaneous emulsifi- interface due to the interaction of free radicals with unsaturated
cation (low energy); MF = Microfluidization (high energy); +S = surfactant was
lipids within the droplets (Arab-Tehrany et al., 2012;
added after fabrication. Numbers in parentheses indicate the final SOR of the
McClements and Decker, 2000). In these systems, lipid oxidation
is usually catalyzed by transition metals, such as iron, which are
also responsible for accelerating lipid oxidation by decomposing
The f-potential was measured throughout emulsion storage to hydroperoxides into free radicals (McClements and Decker, 2000;
provide some indirect information about changes in interfacial Nuchi et al., 2001). The rate of lipid oxidation in emulsions is there-
properties (data not shown). The charges of the emulsions were fore dependent on the relative location of lipid substrates (polyun-
initially between around 1.0 and 1.5 mV, which can be attribu- saturated fatty acids) and pro-oxidants (transition metals and
ted to the fact that the emulsions were stabilized by a non-ionic hydroperoxides) in the system (McClements and Decker, 2000).
surfactant, and that any free fatty acid impurities will be proto- In the current study, lipid hydroperoxide levels and TBARS were
nated under acidic conditions. Haahr and Jacobsen (2008) reported measured to monitor the primary and secondary oxidation prod-
similar values for emulsions prepared from medium chain triglyc- ucts of the emulsions throughout a 14-day oxidation study.
erides, fish oil, Tween 80, and buffer solution (10 mM acetate-im-
idazole buffer, pH 3). The droplet charge of the emulsions Influence of initial particle size. In this series of experiments,
prepared in our study became slightly more negative after 14 days we compared the oxidation in emulsions that had similar
R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20 17

Fig. 6. Microscopic images of emulsions made by MF and SE methods and held at 55 °C with added iron at (A) day 0 and (B) day 14. (C) shows enlarged sections of
abnormalities found in the emulsions on day 14. All emulsions were diluted to 1 wt% total oil phase. Images were taken at 60 magnification. The black bars at the bottom of
each picture are the scales for 10 lm. Key: SE = spontaneous emulsification (low energy); MF = Microfluidization (high energy); +S = surfactant was added after fabrication.
Numbers in parentheses indicate the final SOR of the emulsion.
18 R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20

surfactant levels (SOR = 1.75), but different initial particle sizes and type did (Hu et al., 2003). Studies with fish oil-in-water emulsions
preparation methods. Overall, the time-dependence of lipid oxida- reported that oxidation occurred more quickly in smaller droplets,
tion in all of the emulsions was fairly similar, irrespective of the but this effect could also have been due to the fact that different
preparation method, with all of the emulsions reaching a peak in emulsifier types were used (Haahr and Jacobsen, 2008). Another
hydroperoxide levels after 12 days storage (Fig. 7a). However, the study using fish oil-in-water emulsions also reported that particle
low-energy emulsions with added surfactant did have a slightly size was not a major factor influencing the rate of oxidation, when
higher hydroperoxide value than the other emulsions toward the compared to other factors such as emulsifier type (Sørensen et al.,
end of the incubation period. Interestingly, the SE emulsion with 2008). Overall, these findings suggest that droplet surface area is
added surfactant (SE+S (1.75)) also had a greater particle size not a major factor impacting the chemical stability of encapsulated
(236 nm) than the SE emulsion with SOR 1.75 (SE (1.75); polyunsaturated lipids.
192 nm) and the MF emulsion with added surfactant to SOR 1.75 As for secondary oxidation products, the MF emulsion with
(MF+S (1.75); 212 nm) at these storage times. We had expected a added surfactant to SOR 1.75 (MF+S (1.75)) had the largest initial
higher rate of oxidation in emulsions with smaller particle sizes particle size (166 nm) but ended with an intermediate particle size
because of the greater surface area of the oil droplets, which would (226 nm), reached the highest TBARS value within the 14-day
allow for more oxidation reactions to occur at the droplet surface. study compared to both SE emulsions with SOR 1.75 (SE (1.75)
However, it is also possible for participants in the lipid oxidation and SE+S (1.75)) (Fig. 8a). Based on these results, the fabrication
reaction to partition into the surfactant micelles, which would be method may have a larger impact on the secondary oxidation
expected to alter the lipid oxidation rate. In addition, there may products than the emulsion particle size. The lower concentration
have been only a limited number of reactants in the system and of secondary oxidation products in the low-energy emulsions sug-
all of them may already have been at the droplet surface gests a slower lipid oxidation rate or inhibition of hydroperoxide
(McClements and Decker, 2000; Roozen et al., 1994). degradation.
Previous researchers have also reported different dependences
of lipid oxidation on particle size. Studies of oxidation in protein- Influence of surfactant concentration. All low-energy emul-
stabilized corn oil-in-water emulsions found that particle size did sions (SOR 1.00 and SE+S (1.75); data not shown) and high-energy
not have a major impact on their oxidative stability, but emulsifier emulsions (MF (0.10), MF+S (1.00), and MF+S (1.75)) reached

A SE (1.75)
A SE (1.75) SE+S (1.75)
SE+S (1.75)
70 MF+S (1.75)
Emulsion Hydroperoxides (mM)

1.2 MF+S (1.75)



0.8 40

0.6 30

0.4 20

0 0 5 10 15
0 5 10 15 Storage time (days)
Storage time (days)

B 100
1.2 MF (0.10)
B MF (0.10) MF+S (1.00)
MF+S (1.00) MF+S (1.75)
Emulsion Hydroperoxides (mM)

1 MF+S (1.75) 80

0.4 30

0 0
0 5 10 15 0 5 10 15
Storage time (days) Storage Time (Days)

Fig. 7. Effect of (A) initial particle size and (B) surfactant on hydroperoxide values Fig. 8. Effect of (A) initial particle size and (B) surfactant on TBARS for emulsions
for emulsions made by MF and SE methods and held at 55 °C with added iron for made by MF and SE methods and held at 55 °C with added iron for 14 days. All
14 days. All emulsions were diluted to 1 wt% total oil phase. Key: SE = spontaneous emulsions were diluted to 1 wt% total oil phase. Key: SE = spontaneous emulsifi-
emulsification (low energy); MF = Microfluidization (high energy); +S = surfactant cation (low energy); MF = Microfluidization (high energy); +S = surfactant was
was added after fabrication. Numbers in parentheses indicate the final SOR of the added after fabrication. Numbers in parentheses indicate the final SOR of the
emulsion. emulsion.
R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20 19

similar peak hydroperoxide values on day 12 (Fig. 7b). The SE oxidation in the fish oil emulsions. This result suggests that the
emulsions followed a similar TBARS trend but the SE (1.00) had low turbidity (weak light scattering) of nanoemulsions does not
higher TBARS values than the one with added surfactant (SE+S promote lipid oxidation by allowing more UV-light to penetrate
(1.75)) starting on day 2. In addition the MF emulsion with SOR into them. Overall, this study shows that a low-energy homoge-
0.10 (MF (0.10) reached a peak TBARS value on day 14 while the nization method (spontaneous emulsification) can be used to pro-
emulsions with added surfactant (MF+S (1.00) and MF+S (1.75)) duce fish oil emulsions that may be suitable to fortify transparent
had high TBARS values on day 10, which plateaued through day food or beverage systems. The main advantage of this method is
14 (Fig. 8b). that it is simple and rapid and does not require any expensive
It was expected that the higher concentration of surfactant equipment (such as a homogenizer), however, the main disadvan-
would have slowed down the rate of oxidation in the emulsions. tage is that it requires relatively high levels of synthetic
Surfactants affect many characteristics of an emulsion including surfactants.
droplet charge, interfacial layer thickness, and permeability, which
in turn influence accessibility of pro-oxidants, free radicals, and
oxygen to the droplet lipids (Villiere et al., 2005; Waraho et al.,
2011). Surfactant can be found surrounding the oil droplets in
This material is based upon work supported by the Cooperative
the emulsion and as surfactant micelles in the aqueous phase when
State Research, Extension, Education Service, United State
excess surfactant is used (McClements and Decker, 2000). In this
Department of Agriculture, Massachusetts Agricultural
case, the non-ionic surfactant produced minimal particle charge
Experiment Station (Project No. 831) and by the United States
and because of that, it is not expected to have impacted the lipid
Department of Agriculture, NRI Grants (2011-03539, 2013-03795,
oxidation greatly. Surfactant at the interfacial layer can also form
2011-67021, and 2014-67021).
a physical (steric) barrier between pro-oxidants and the lipid phase
(Silvestre et al., 2000; Waraho et al., 2011). The effectiveness of the
interfacial layer relies on its thickness, which is dependent on sur- References
factant’s head group. Brij 700 has a larger head group than Brij 76
and was found to be more effective at decreasing the rate of oxida- Anton, N., Vandamme, T.F., 2009. The universality of low-energy nano-
emulsification. Int. J. Pharm. 377 (1–2), 142–147.
tion than Brij 76 (Silvestre et al., 2000). Anton, N., Vandamme, T.F., 2011. Nano-emulsions and micro-emulsions:
Tween 80 has a low critical micelle concentration (CMC; clarifications of the critical differences. Pharm. Res. 28 (5), 978–985.
<0.1 mM), a point where the surface of the lipid particles are satu- Anton, N., Benoit, J.-P., Saulnier, P., 2008. Design and production of nanoparticles
formulated from nano-emulsion templates—a review. J. Control. Release 128
rated with surfactant and the excess surfactant forms surfactant
(3), 185–199.
micelles (Patist et al., 2000). Surfactant micelles have been shown Arab-Tehrany, E., Jacquot, M., Gaiani, C., Imran, M., Desobry, S., Linder, M., 2012.
to decrease lipid oxidation through multiple mechanisms Beneficial effects and oxidative stability of omega-3 long-chain polyunsaturated
(Richards et al., 2002). One such way is by solubilizing iron or fatty acids. Trends Food Sci. Technol. 25 (1), 24–33.
Binks, B.P., Clint, J.H., Fletcher, P.D.I., Rippon, S., Lubetkin, S.D., Mulqueen, P.J., 1999.
hydroperoxides as a means to decrease pro-oxidants in the emul- Kinetics of swelling of oil-in-water emulsions stabilized by different
sion (Cho et al., 2002; Nuchi et al., 2002). Secondly, surfactants, surfactants. Langmuir 15 (13), 4495–4501.
including Tween 20, have been shown to increase the solubiliza- Capek, I., 2004. Degradation of kinetically-stable o/w emulsions. Adv. Colloid
Interface Sci. 107 (2–3), 125–155.
tion of antioxidants into the aqueous phase of the emulsion Cho, Y.-J., McClements, D.J., Decker, E.A., 2002. Ability of surfactant micelles to alter
(Kiralan et al., 2014). Lastly, it has also been hypothesized that the physical location and reactivity of iron in oil-in-water emulsion. J. Agric.
tocopherols in fish oil can form surfactant co-micelles, which act Food Chem. 50 (20), 5704–5710.
Fanun, M., 2008. Microemulsions: Properties and Applications. CRC Press, Boca
as a reservoir for the antioxidants and they can be released as they Raton, FL.
were needed to extend the lag period of the oxidation process Glanz, K., Basil, M., Maibach, E., Goldberg, J., Snyder, D.A.N., 1998. Why Americans
(Kiralan et al., 2014). eat what they do: taste, nutrition, cost, convenience, and weight control
concerns as influences on food consumption. J. Am. Diet. Assoc. 98 (10), 1118–
Despite all of the potentially positive impacts that surfactant 1126.
can have on lipid oxidation in emulsions, surfactant itself can form Gulotta, A., Saberi, A.H., Nicoli, M.C., McClements, D.J., 2014a. Nanoemulsion-based
hydroperoxides (Nuchi et al., 2001). The polyethers found in the delivery systems for polyunsaturated (omega-3) oils: formation using a
spontaneous emulsification method. J. Agric. Food Chem. 62 (7), 1720–1725.
hydrophilic head group of Tweens are easily oxidized and form
Gulotta, A., Saberi, A.H., Nicoli, M.C., McClements, D.J., 2014b. Nanoemulsion-based
hydroperoxides and their degradation products (Lever, 1977; delivery systems for polyunsaturated (x-3) oils: formation using a spontaneous
Nuchi et al., 2001; Ray and Puvathingal, 1985). Nuchi et al. showed emulsification method. J. Agric. Food Chem. 62 (7), 1720–1725.
that Tween 20 high in hydroperoxides increased the rate of oxida- Haahr, A.-M., Jacobsen, C., 2008. Emulsifier type, metal chelation and pH affect
oxidative stability of n-3-enriched emulsions. Eur. J. Lipid Sci. Technol. 110 (10),
tion in salmon oil emulsions (2001). The oxidation of the surfac- 949–961.
tants themselves may have contributed to the hydroperoxides Hu, M., McClements, D.J., Decker, E.A., 2003. Lipid oxidation in corn oil-in-water
measured in the fish oil emulsions in our study instead of slowing emulsions stabilized by casein, whey protein isolate, and soy protein isolate. J.
Agric. Food Chem. 51 (6), 1696–1700.
down the oxidation. Additionally, the high temperature and addi- Jafari, S.M., He, Y.H., Bhandari, B., 2006. Nano-emulsion production by sonication
tion of iron into the fish emulsions may have increase the oxidation and microfluidization – a comparison. Int. J. Food Prop. 9 (3), 475–485.
rate making it difficult to find differences in the oxidation rate Jafari, S.M., He, Y., Bhandari, B., 2007. Optimization of nano-emulsions production
by microfluidization. Eur. Food Res. Technol. = Zeitschrift für Lebensmittel-
between the emulsions. Untersuchung und -Forschung. A 225 (5-6), 733–741.
Jafari, S.M., Assadpoor, E., He, Y., Bhandari, B., 2008. Re-coalescence of emulsion
5. Conclusions droplets during high-energy emulsification. Food Hydrocolloids 22 (7), 1191–
Kabalnov, A., 2001. Ostwald ripening and related phenomena. J. Dispersion Sci.
This study determined that the low-energy method of sponta- Technol. 22 (1).
neous emulsification can produce optically transparent nanoemul- Kabalnov, A.S., Shchukin, E.D., 1992. Ostwald ripening theory: applications to
fluorocarbon emulsion stability. Adv. Colloid Interface Sci. 38, 69–97.
sions at higher surfactant-to-oil ratios (SOR P 1.25). These
Kennedy, E.T., Luo, H., Ausman, L.M., 2012. Cost implications of alternative sources
nanoemulsions were physically stable when stored at 37 °C for of (n-3) fatty acid consumption in the United States. J. Nutr. 142 (3), 605S–609S.
14 days, but exhibited some droplet growth when stored at a Kiralan, S.S., Doğu-Baykut, E., Kittipongpittaya, K., McClements, D.J., Decker, E.A.,
higher temperature, which was attributed to coalescence caused 2014. Increased antioxidant efficacy of tocopherols by surfactant solubilization
in oil-in-water emulsions. J. Agric. Food Chem. 62 (43), 10561–10566.
by surfactant head group dehydration. Neither particle size nor Kris-Etherton, P.M., Grieger, J.A., Etherton, T.D., 2009. Dietary reference intakes for
surfactant concentration had a major impact on the rate of lipid DHA and EPA. Prostaglandins Leukot. Essent. Fatty Acids 81 (2–3), 99–104.
20 R.M. Walker et al. / Journal of Food Engineering 164 (2015) 10–20

Lever, M., 1977. Peroxides in detergents as interfering factors in biochemical Racine, R.A., Deckelbaum, R.J., 2007. Sources of the very-long-chain unsaturated
analysis. Anal. Biochem. 83 (1), 274–284. omega-3 fatty acids: eicosapentaenoic acid and docosahexaenoic acid. Curr.
Maki, K.C., Yurko-Mauro, K., Dicklin, M.R., Schild, A.L., Geohas, J.G., 2014. A new, Opin. Clin. Nutr. Metab. Care 10 (2), 123–128.
microalgal DHA- and EPA-containing oil lowers triacylglycerols in adults with Rao, J., McClements, D.J., 2011. Formation of flavor oil microemulsions,
mild-to-moderate hypertriglyceridemia. Prostaglandins, Leukotrienes Essential nanoemulsions and emulsions: influence of composition and preparation
Fatty Acids (PLEFA) 91 (4), 141–148. method. J. Agric. Food Chem. 59 (9), 5026–5035.
Mason, T.G., Wilking, J.N., Meleson, K., Chang, C.B., Graves, S.M., 2006. Rao, J., McClements, D.J., 2013. Optimization of lipid nanoparticle formation for
Nanoemulsions: formation, structure, and physical properties. J. Phys.: beverage applications: influence of oil type, cosolvents, and cosurfactants on
Condens. Matter 18 (41). nanoemulsion properties. J. Food Eng. 118 (2), 198–204.
Mayer, S., Weiss, J., McClements, D.J., 2013. Vitamin E-enriched nanoemulsions Ray Jr., W.J., Puvathingal, J.M., 1985. A simple procedure for removing
formed by emulsion phase inversion: factors influencing droplet size and contaminating aldehydes and peroxides from aqueous solutions of
stability. J. Colloid Interface Sci. 402, 122–130. polyethylene glycols and of nonionic detergents that are based on the
McClements, D.J., 1994. Ultrasonic determination of depletion flocculation in oil-in- polyoxyethylene linkage. Anal. Biochem. 146 (2), 307–312.
water emulsions containing a nonionic surfactant. Colloids Surf. A- Richards, M.P., Chaiyasit, W., McClements, D.J., Decker, E.A., 2002. Ability of
Physicochem. Eng. Aspects 90 (1), 25–35. surfactant micelles to alter the partitioning of phenolic antioxidants in oil-in-
McClements, D.J., 2002. Theoretical prediction of emulsion color. Adv. Colloid water emulsions. J. Agric. Food Chem. 50 (5), 1254–1259.
Interface Sci. 97 (1–3), 63–89. Roozen, J.P., Frankel, E.N., Kinsella, J.E., 1994. Enzymic and autoxidation of lipids in
McClements, D.J., 2011. Edible nanoemulsions: fabrication, properties, and low fat foods: model of linoleic acid in emulsified hexadecane. Food Chem. 50
functional performance. Soft Matter 7 (6), 2297–2316. (1), 33–38.
McClements, D.J., 2012. Nanoemulsions versus microemulsions: terminology, Saberi, A.H., Fang, Y., McClements, D.J., 2013. Fabrication of vitamin E-enriched
differences, and similarities. Soft Matter 8 (6), 1719–1729. nanoemulsions: factors affecting particle size using spontaneous emulsification.
McClements, D.J., Decker, E.A., 2000. Lipid oxidation in oil-in-water emulsions: J. Colloid Interface Sci. 391, 95–102.
impact of molecular environment on chemical reactions in heterogeneous food Shantha, N.C., Decker, E.A., 1994. Rapid, sensitive, iron-based spectrophotometric
systems. J. Food Sci. 65 (8), 1270–1282. methods for determination of peroxide values of food lipids. J. AOAC Int. 77 (2),
McClements, D.J., Rao, J., 2011. Food-grade nanoemulsions: formulation, fabrication, 421–424.
properties, performance, biological fate, and potential toxicity. Crit. Rev. Food Silvestre, M.P.C., Chaiyasit, W., Brannan, R.G., McClements, D.J., Decker, E.A., 2000.
Sci. Nutr. 51 (4), 285–330. Ability of surfactant headgroup size to alter lipid and antioxidant oxidation in
McClements, D.J., Decker, E.A., Weiss, J., 2007. Emulsion-based delivery systems for oil-in-water emulsions. J. Agric. Food Chem. 48 (6), 2057–2061.
lipophilic bioactive components. J. Food Sci. 72 (8), R109–R124. Solans, C., Sole, I., 2012. Nano-emulsions: formation by low-energy methods. Curr.
McDonald, R.E., Hultin, H.O., 1987. Some characteristics of the enzymic lipid Opin. Colloid Interface Sci. 17 (5), 246–254.
peroxidation system in the microsomal fraction of flounder skeletal muscle. J. Sørensen, A.-D.M., Haahr, A.-M., Becker, E.M., Skibsted, L.H., Bergenståhl, B., Nilsson,
Food Sci. 52 (1), 15–21. L., Jacobsen, C., 2008. Interactions between iron, phenolic compounds,
Nuchi, C.D., McClements, D.J., Decker, E.A., 2001. Impact of Tween 20 emulsifiers, and pH in omega-3-enriched oil-in-water emulsions. J. Agric.
hydroperoxides and iron on the oxidation of methyl linoleate and salmon oil Food Chem. 56 (5), 1740–1750.
dispersions. J. Agric. Food Chem. 49 (10), 4912–4916. Tadros, T., Izquierdo, P., Esquena, J., Solans, C., 2004. Formation and stability of
Nuchi, C.D., Hernandez, P., McClements, D.J., Decker, E.A., 2002. Ability of lipid nano-emulsions. Adv. Colloid Interface Sci. 108–109, 303–318.
hydroperoxides to partition into surfactant micelles and alter lipid oxidation Villiere, A., Viau, M., Bronnec, I., Moreau, N., Genot, C., 2005. Oxidative stability of
rates in emulsions. J. Agric. Food Chem. 50 (19), 5445–5449. bovine serum albumin- and sodium caseinate-stabilized emulsions depends on
Ostertag, F., Weiss, J., McClements, D.J., 2012. Low-energy formation of edible metal availability. J. Agric. Food Chem. 53 (5), 1514–1520.
nanoemulsions: factors influencing droplet size produced by emulsion phase Waraho, T., McClements, D.J., Decker, E.A., 2011. Mechanisms of lipid oxidation in
inversion. J. Colloid Interface Sci. 388, 95–102. food dispersions. Trends Food Sci. Technol. 22 (1), 3–13.
Patist, A., Bhagwat, S.S., Penfield, K.W., Aikens, P., Shah, D.O., 2000. On the Yang, Y., Marshall-Breton, C., Leser, M.E., Sher, A.A., McClements, D.J., 2012.
measurement of critical micelle concentrations of pure and technical-grade Fabrication of ultrafine edible emulsions: comparison of high-energy and
nonionic surfactants. J. Surfactants Deterg. 3 (1), 53–58. low-energy homogenization methods. Food Hydrocolloids 29 (2), 398–406.