Food Hydrocolloids 81 (2018) 120e128

Contents lists available at ScienceDirect

Food Hydrocolloids
journal homepage: www.elsevier.com/locate/foodhyd

Improved emulsion stability and modified nutrient release by

structuring O/W emulsions using konjac glucomannan
Wei Lu a, b, Baodong Zheng c, Song Miao a, c, *
a
Teagasc Food Research Centre, Moorepark, Fermoy, Cork, Ireland
b
School of Food and Nutritional Sciences, University College Cork, Cork, Ireland
c
China-Ireland International Cooperation Centre for Food Material Science and Structure Design, Fujian Agriculture and Forestry University, Fuzhou, China

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

Article history: Functional konjac glucomannan (KGM) was used to structure the water phase of O/W emulsions con-
Received 13 November 2017 taining a lipophilic bioactive compound (b-carotene). KGM greatly increased the viscosity of the water
Received in revised form phase and thus the viscosity of final emulsions. Results of Fourier-transform infrared spectroscopy (FT-IR)
24 January 2018
showed that there is no significant non-covalent interaction between KGM and whey proteins in the
Accepted 19 February 2018
Available online 22 February 2018
water phase. KGM significantly improved the creaming and pH stability of whey-protein-stabilized
emulsions (p < 0.05), and significantly decreased the oiling-off of emulsions during freeze-thaw test.
Emulsions with or without KGM all had good thermal stability at 80
C. Microscopy observations indi-
Keywords:
Structuring emulsions
cated obvious aggregation of free proteins and oil droplets in gastric phase and an enzymatic-induced
Konjac glucomannan break-down of droplets, mainly in the intestinal phase of the simulated gastrointestinal tract (GIT)
Emulsion stability digestion. Emulsions with KGM-structured water phase showed a lower final release rate of encapsulated
Digestion b-carotene than emulsion without KGM (p < 0.05), and the release rate decreased with the increasing
Nutrient release KGM content. The findings of this study contribute to a better understanding of the influence of the water
phase on the release of encapsulated compounds from emulsions, and make it possible to achieve
controlled release of encapsulated compounds, and/or to deliver multiple health-beneficial nutrients at
once by structuring emulsion-based carriers with functional natural biopolymers.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction structure, including the oil phase, interfacial layer, and water phase.
Taking protein stabilized oil-in-water (O/W) emulsion as an
There has been a growing interest in the utilization of emulsions example, the digestion process of protein-stabilized emulsion
as the carriers for the encapsulation, protection and delivery of droplets containing bioactive nutrients mainly goes through four
lipophilic functional nutrients. Many emulsion-based carriers have steps: (i) binding of proteinase to the droplet surface in the water
been successfully developed to protectively deliver a variety of phase, e.g., pepsin and trypsin; (ii) hydrolysis of interfacial protein
bioactive nutrients, such as carotenoids (Lu, Kelly, & Miao, 2017b; layers by proteinase; (iii) hydrolysis of lipid in the oil phase by
Lu et al., 2016), fatty acids (Karthik & Anandharamakrishnan, 2016), lipase and release of bioactive nutrients from lipid droplets; and
polyphenols (Lu, Kelly, & Miao, 2016), and peptides (Niu, Conejos- (iv) formation of bile-salt emulsified micelles containing released
Sanchez, Griffin, O'Driscoll, & Alonso, 2016). lipophilic bioactive nutrients. Accordingly, the formula of water
Since emulsions have been widely used as novel carriers for phase and interface can influence the final bioaccessibility by
functional nutrients, a better understanding of the factors that can modifying the first two steps of the digestion, while the structure
influence the digestion behavior of emulsion droplets and thus the and composition of oil phase can influence the final bioaccessibility
release of encapsulated bioactive nutrients becomes very impor- through affecting the later two steps of digestion. The effects of
tant. The digestion of emulsion droplets is closely related to their composition or structure of the oil phase and interface on the
release of encapsulated bioactive nutrients from emulsions have
been well investigated in previous studies (Lu et al., 2017b; Qian,
Decker, Xiao, & McClements, 2012; Salvia-Trujillo, Fumiaki, Park,
* Corresponding author. Teagasc Food Research Centre, Moorepark, Fermoy, Cork,
Ireland. & McClements, 2017). However, little is known about the influ-
E-mail address: [email protected] (S. Miao). ence of the water phase on the digestion of emulsion droplets and

https://doi.org/10.1016/j.foodhyd.2018.02.034
0268-005X/© 2018 Elsevier Ltd. All rights reserved.
 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128 121

thus the release of encapsulated nutrients from emulsions. as the water phase. The oil phase was prepared by dissolving b-
Natural polysaccharides are a class of biopolymers that have carotene (0.2%, w/w in final emulsion) in sunflower oil (10%, w/w in
been widely used in the production of emulsions, due to their wide final emulsion) at 140
C which was then mixed with the water
availability, good physical and chemical stability, edibility, and low phase at 10,000 rpm for 2 min at room temperature using an Ultra-
cost. Previous studies have shown that introducing polysaccharide Turrax (IKA, Staufen, Germany) followed by further homogeniza-
as a second stabilizer into the water phase can significantly tion (APV 1000, SPX Flow Technology, Charlotte, North Carolina,
enhance the stability of emulsions (Dickinson, 2011). Poly- USA) at 50 MPa for 3 passes at room temperature to obtain final
saccharides can generally modify the interface, rheology, or gela- emulsions.
tion properties of emulsions. Polysaccharides can also significantly
enhance the stability of protein-stabilized emulsions by forming a 2.3. Droplet size and surface charge
polysaccharide-protein double-layer interface (Aoki, Decker, &
McClements, 2005; Guzey, Kim, & McClements, 2004; Jeonghee The droplet size and zeta potential of KGM emulsions were
Surh, Decker, Julian McClements, 2005; Mao, Roos, & Miao, 2015). measured by a laser particle analyzer (Nano-ZS, Malvern In-
In addition, polysaccharides can form network structures in the struments, Worcestershire, UK) as described in our previous study
water phase, which can limit the mobility of oil droplets by steric (Lu et al., 2016). Emulsions were diluted to the final oil content (w/
hindrance and thus improve the creaming stability of emulsions w) of 0.01% before testing. The refractive index (RI) of samples was
(Lin et al., 2017). Such network structures in the water phase can set at 1.47 for sunflower oil.
potentially also influence the binding and interaction process of
enzymes, the digestion velocity of emulsion droplets, and thus the 2.4. Creaming stability
release of encapsulated bioactive nutrients, as described above.
Konjac glucomannan (KGM) is a natural polysaccharide ob- The creaming stability of emulsions was evaluated using
tained from tubers of Amorphophallus konjac cultivated in Asia. Lumisizer (LUM GmbH, Berlin, Germany) as describe previously.
KGM is reported to possess many health benefits, such as lowering Emulsions were centrifuged at 2300 g at 25
C with a scanning rate
the blood cholesterol and sugar levels, positively modulating gut of once every 10 s for 1200 s. Following the test, curves of the in-
microflora, promoting weight loss, and improving immune func- tegrated transmitted light against time were plotted, and the slope
tion (Arvill and Lennart, 1995; Chua, Baldwin, Hocking, & Chan, of each curve was taken as the Creaming Index (CI).
2010). However, there are few studies on using KGM to modify
the water phase and thus influencing the release of encapsulated 2.5. Rheological analysis
bioactive nutrients from emulsions. Introducing KGM into emul-
sions can also endow the final emulsions with new health benefits Rheological measurements were performed using an AR 2000ex
associated with KGM besides those contributed by encapsulated rheometer (TA Instruments, Crawley, UK). A concentric cylinder
molecules. This can accordingly achieve a delivery of multiple nu- geometry (stator inner radius ¼ 15 mm, rotor outer
trients in one carrier. radius ¼ 14 mm) was selected, and 19 g of each sample was placed
From the above, this study was therefore proposed to investi- into the inner cylinder and equilibrated for 2 min before mea-
gate the effect of structuring water phase of model O/W emulsions surement. Viscosity testing was performed over a shear rate range
by KGM on their properties, including droplet size, surface charge, of 0e300 s1 at 25
C.
creaming stability, pH stability, thermal stability, and freeze-thaw
stability, and also the effect on the release of encapsulated bioac- 2.6. Fourier transforms infrared spectroscopy (FT-IR)
tive nutrient (b-carotene) from KGM-structured O/W emulsions
after passing through a simulated gastrointestinal tract (GIT) Fourier transform infrared spectroscopy (FT-IR) technology was
digestion. used to evaluate the molecular structure of KGM and WPI in water
phase before and after homogenization. The IR spectra of the
2. Material and methods samples were recorded by FT-IR spectrophotometer (Bruker Corp,
Billerica, Massachusetts, USA) using an attenuated total reflection
2.1. Materials (ATR) technique. The spectrum was scanned from 4000 cm1 to
900 cm1 with a resolution of 4 cm1. An average of 300 scans was
Konjac glucomannan powder was obtained from Konjac Food recorded for each sample.
(Cupertino, CA, USA). Whey protein isolate was purchased from
Davisco Food International (Le Sueur, MN, USA). b-carotene (>93%, 2.7. Stability of emulsions at different pH values and temperature
UV), pepsin, pancreatin (porcine, 4  USP) were purchased from
Sigma-Aldrich (St. Louis, MO, USA). Sunflower oil was purchased The effect of KGM on the resistance of emulsions to extreme pH
from a local supermarket. All other chemicals and reagents used environments was evaluated. The KGM emulsions were brought to
were of AR-grade and obtained from Sigma-Aldrich (St. Louis, MO, different pH values from 2.0 to 7.0 with HCI or NaOH solution, and
USA). maintained for 4 h at room temperature before droplet size, surface
charge and creaming stability analysis by DLS and Lumisizer as
2.2. Preparation of emulsions described above.
The effect of KGM on the resistance of emulsions to thermal
Whey protein isolate (WPI) was dispersed (2%, w/w) in Millipore processing was evaluated following incubation at 25
C, 37
C, or
water containing sodium azide as antimicrobial agent (0.01% w/w). 80
C for 2 h. The droplet size was then analyzed by Malvern
The dispersion was stirred for 4 h and kept at 4
C overnight for Nanosizer as described above.
complete dissolution of WPI. The dispersion was then brought to
25
C before adding konjac glucomannan (KGM) powder to make a 2.8. Freeze-thaw test
KGM content of 0.05%, 0.1%, or 0.2% in the final emulsions. The
mixtures were stirred for 4 h for a complete dissolution of KGM and Freeze-thaw testing was applied to emulsions with the objective
then centrifuged at 4000 rpm (2700 g) for 20 min before being used of assessing the possibility of frozen storage of liquid emulsions
122 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128

containing bioactive nutrients. In addition, the changes in proper- phase of GIT was measured as the final release rate of encapsulated
ties of emulsions after the freeze-thaw processing can potentially b-carotene. Briefly, an aliquot of raw digesta after the intestinal
contribute to the freeze-drying of liquid emulsions into solid phase digestion was centrifuged at 4500 rpm (2978 g) for 40 min at
products, which can significantly facilitate the storage, trans- 4
C and the middle layer was collected and considered as the
portation, and application of emulsions. micelle fraction. One mL of the micelle fraction was extracted twice
Liquid emulsions were kept at 20
C for 24 h and then thawed with ethanol/n-hexane. The upper n-hexane layer containing the
at 25
C for 2 h in a water bath. This cycle was repeated three times. solubilized b-carotene was collected and analyzed by RP-HPLC as
Droplet size and surface charge was measured after each cycle by described below.
Malvern nanosizer. Creaming stability of emulsions after three cy- The final release rate of encapsulated b-carotene was calculated
cles was measured with Lumisizer as described above. using the follow equation:
Oiling-off of emulsions after 3 cycles of freeze-thaw was also
evaluated by measuring the content of b-carotene in the free oil Cmicelle
Release rate ð%Þ  100% (2)
fraction. The thawed emulsions were centrifuged at 10,000 g for Cinitial
10 min (25
C). Free oil layer containing b-carotene on top of
emulsions were collected and subjected to a second centrifuge at where Cmicelle and Cinitial are the concentration of b-carotene in the
4000 g for 10 min (25
C). The supernatant (free oil) containing b- micelle fraction and initial emulsion before digestion, respectively.
carotene was collected and extracted with ethanol/hexane. The
content of b-carotene in hexane fraction was quantified by RP-HPLC 2.12. Quantification of b-carotene
as described below. The oiling-off was calculated based on the
equation below: Reversed-phase high performance liquid chromatography (RP-
HPLC) was used to quantify b-carotene. An Agilent 1200 series
Cfree oil system with a DAD UV-Vis detector (Agilent, Santa Clara, CA, USA)
Oil  off ð%Þ ¼  100% (1) and a reversed phase TSKgel ODS-100v C18 column (4.6  250 mm,
Cinitial
5 mm, TOSOH) was employed.
where Cfree oil and Cinitial are the concentration of b-carotene in the Chromatography conditions: column operation temperature at
free oil fraction after 3 cycles of freeze-thaw and in the initial 30
C; elution was performed with 90% ethanol and 10% acetonitrile
emulsion before freeze-thaw test, respectively. from 0 to 30 min, flow rate was 1 mL/min, detection wavelength
was 450 nm, and injection volume was 20 mL.
2.9. Simulated gastrointestinal tract (GIT) digestion
2.13. Statistical analysis
An in vitro simulated GIT model consisting of mouth, gastric and
intestinal phases was used to digest b-carotene loaded emulsions. All experiments were repeated at least three times. One-way
The simulated saliva fluid (SSF), simulated gastric fluid (SGF), and analysis of variance (ANOVA) was employed to compare means of
simulated intestinal fluid (SIF) were prepared as described previ- data. A t-Test was used to determine the differences between
ously (Lu et al., 2017b) with minor modification. means. Significant differences were determined at the 0.05 level
For the mouth phase digestion, emulsions were mixed with SSF (p < 0.05).
(1:1, v/v), the pH was adjusted to 6.8 and the mixtures were
incubated at 37
C for 10 min with continuous agitation at 3. Results and discussion
150 rpm.For the gastric phase, the bolus sample from the mouth
phase was mixed with the SGF (1:1, v/v). The pH of the mixture was 3.1. Droplet size and surface charge
adjusted to 2.5 and it was incubated at 37
C for 2 h with contin-
uous agitation at 150 rpm. The pepsin activity in the final mixture As is shown in Table 1, an emulsion without konjac gluco-
was 1000 U/ml. For the small intestinal phase, the bolus sample mannan (KGM) had an average droplet size of 252 nm, while
from the gastric phase was mixed with the SIF (1:1, v/v). The pH of emulsions containing KGM showed significantly increased droplet
the mixture was adjusted to 7.0 and it was incubated at 37
C for 2 h size with increasing KGM content from 0.05% to 0.2% (p < 0.05).
with continuous agitation at 150 rpm. The concentration of KGM, as a natural polysaccharide, is dispersible in water and can
pancreatin (porcine, 4  USP) in the final mixture was based on the form a network structure with different particle size depending on
trypsin activity (25 U/ml). its molecular weight and concentration (Lazaridou, Biliaderis, &
Izydorczyk, 2003). This network structure can potentially increase
2.10. Laser scanning confocal microscope observation the average droplet size of emulsions. However, the droplet size
test was performed after 1000-fold dilution of emulsions and KGM
The microstructure of the initial and digested emulsion droplets at this extreme low concentration (<0.0002%) is unlikely to form a
was observed using a Leica TCS SP5 laser scanning confocal mi- network structure.
croscope (Leica Microsystems, Baden-Württemberg, Germany). All
the images were taken using a 63 x oil-immersion objective and
Table 1
simultaneous dual-channel imaging, He-Ne laser (excitation Droplet size, zeta potential, polydispersity index (PdI) and creaming index (CI) of
wavelength at 633 nm) and an Argon laser (excitation wavelength emulsions containing KGM.
at 488 nm). A mixture of two dyes, Fast green (0.1%, w/w in water)
KGM (%, w/w) size (d.nm) zeta potential (mV) PdI CI (%/min)
and Nile red (0.1%, w/w in propanediol) was used to color and
detect protein and lipid, respectively. Initial/digested sample 0 252±9d 65.0 ± 5.4a 0.24 ± 0.02a 0.64 ± 0.01b
0.05 280±8c 60.4 ± 2.2a 0.21 ± 0.02b 0.52 ± 0.02c
(500 ml) was gently mixed with 50 ml of mixed dye. 0.1 306 ± 17b 59.3 ± 2.6a 0.21 ± 0.01b 0.55 ± 0.02c
0.2 350 ± 24a 59.3 ± 0.7a 0.20 ± 0.02b 1.70 ± 0.04a
2.11. Release of encapsulated b-carotene *KGM indicates konjac glucomannan.
a
Different letters indicate significant difference between values in a column
The amount of b-carotene in micelle fractions after the intestinal (p < 0.05).
 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128 123

close enough at high KGM concentrations induced by depletion

attraction between droplets (Ye, Hemar, & Singh, 2004). Slight
aggregation of droplets was also observed in our previous research
(Lu et al., 2016).
All emulsions were negatively charged and no significant dif-
ference in surface charge between different emulsions was
observed (p > 0.05). Generally, KGM is a non-charged poly-
saccharide and binding of KGM to the surface of WPI-stabilized
emulsion droplets should lead to their significantly reduced sur-
face charge. However, surface charge of emulsions containing KGM
showed almost no difference from that of the emulsion without
KGM (Table 1), potentially demonstrating that there is very limited
binding of KGM to the droplet surface.

3.2. Rheological analysis

The viscosity of emulsions significantly decreased with the

increasing shear rate (~50 1/s), indicating a shear-thinning prop-
Fig. 1. Viscosity of emulsions. KGM 0% indicates emulsion without konjac gluco-
mannan (KGM), while KGM 0.05%, KGM 0.1%, and KGM 0.2% indicates emulsions with erty. The KGM emulsion showed higher viscosity than emulsion
KGM content (w/w) of 0.05%, 0.1%, and 0.2%, respectively. Insert: initial viscosity at very without KGM, and the viscosity of KGM emulsions increased with
low shear rate (~2.5 1/s). increasing KGM content (Fig. 1). Several factors can potentially in-
fluence the viscosity of emulsions, such as oil content, viscosity of
water phase, droplet size, or surface charge (McClements, 2015). In
There is another possibility to explain the increased droplet size this study, increased viscosity of KGM emulsions can be mainly
of KGM-emulsions. The peak of the size distribution shifted to a attributed to two factors: (i) increased viscosity of water phase
slightly higher value (data not shown), indicating that there may be (WPI-KGM dispersion); and (ii) flocculation of droplets by deple-
some flocculation or coalescence of the droplets, which were forced tion force induced by non-absorbed polysaccharide KGM. As shown

Fig. 2. Viscosity of whey protein isolate (WPI) and konjac glucomannan (KGM) mixed solutions before and after homogenization at 50 MPa. (a) WPI solution without KGM; (b) WPI
solution with 0.05% (w/w) KGM; (c) WPI solution with 0.1% KGM; (d) WPI solution with 0.2% KGM.
124 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128

in Fig. 2, the viscosity of the water phase (WPI-KGM dispersions) preliminary research also indicated that there is no significant
also decreased with increasing shear rate, and a similar increase in hydrophobic or hydrogen bond interaction between WPI and KGM
viscosity with increasing KGM content was seen, suggesting that at different pH values (2e8) and temperatures (up to 90
C) (data
KGM can significantly increase the viscosity of the water phase and not shown).
thus the viscosity of final emulsions. In addition, KGM, as a However, little was known about the interaction between WPI
biopolymer can potentially generate a depletion force between and KGM in the water phase after homogenization. Hence, fourier-
droplets and induce flocculation of droplets at certain concentra- transform infrared spectroscopy (FT-IR) was employed to investi-
tions (Mao et al., 1995), which accordingly can increase the vis- gate the influence of homogenization process on their structures
cosity of emulsions. This point will be further discussed in the and potential interactions. The spectra of WPI-KGM dispersions
creaming stability section below. before and after homogenization were collected within wave-
Generally, KGM can disperse in water and form highly viscous lengths of 900e4000 cm1 and spectra between 1350 and
suspensions at pH values of 4.0e7.0 due to its high molecular 1700 cm1 were analyzed, since this zone covers the main char-
weight, ranging from 200 to 2000 kDa (Chua et al., 2010; Villay acteristic absorption peaks of WPI (Chen, Li, Ding, & Suo, 2012). The
et al., 2012). The viscosity of WPI-KGM suspensions after homog- spectra of WPI (Fig. 3) indicated the presence of characteristic ab-
enization decreased significantly (Fig. 2). This is mainly attributed sorption peaks at 1645, 1548, 1456, and 1400 cm1 corresponding
to the mechanical de-polymerization and/or de-polymerization- to the C¼O stretching and the bending of N-H, C-H, C-N bonds,
coupled conformation of KGM by homogenization (Villay et al., respectively (Barth, 2007; Chen et al., 2012). Compared with WPI
2012). The molar-mass distribution is the primary parameter that before homogenization, the spectra of WPI after homogenization
influences the viscosity of polysaccharide in solution. Homogeni- did not show significant difference.
zation can lead to mechanical degradation (de-polymerization) of KGM exhibited a characteristic absorption peak of the b-1,4-
polysaccharides, and produce fractions with low molecular weight linked glycosidic bond at 895 cm1 and a characteristic peak of
or low polydispersity, which accordingly lead to decreased the enlargement of pyranoid rings at 808 cm1 (Gao, Su, Huang, &
viscosity. He, 2014). Hence, KGM showed almost no significant absorption
within wavelength of 900-4000 cm1. Spectra of WPI-KGM dis-
3.3. Fourier-transform infrared spectroscopy (FT-IR) analysis persions also showed no significant difference compared with pure

Generally, polysaccharides can form double layer structures at

the interface of protein-stabilized emulsion through electrostatic
attraction, hydrogen bond or hydrophobic interactions with pro-
tein. However, KGM is a non-charged polysaccharide and cannot
form complexes with protein by electrostatic attraction. Our

Fig. 3. Fourier-transform infrared (FT-IR) spectra of (a) WPI (2%, w/w), (b) WPI (2%, w/
w) after homogenization, (c) a mixture of WPI (2%, w/w) and KGM (0.2%, w/w) and (d) Fig. 4. Droplet size (a) and zeta potential (b) of emulsions at different pH values. KGM
a mixture of WPI (2%, w/w) and KGM (0.2%, w/w) after homogenization. WPI indicates 0% indicate emulsion without konjac glucomannan (KGM). KGM 0.05%, KGM 0.1%, and
whey protein isolate. KGM indicates konjac glucomannan. KGM 0.2% indicates emulsions with KGM content of 0.05%, 0.1%, and 0.2%, respectively.
 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128 125

WPI dispersion before and after homogenization (Fig. 3). All these creaming stability, which was mainly attributed to the depletion
results suggested that: (i) homogenization did not induce signifi- flocculation of emulsion droplets by non-absorbed KGM
cant changes in the molecular structure of WPI; and (ii) homoge- (Klinkesorn, Sophanodora, Chinachoti, & McClements, 2004; Mao
nization processing did not induce obvious non-covalent et al., 1995). Non-adsorbed polymers can generate an attractive
interactions, e.g., electrostatic attraction, hydrophobic interaction osmotic force between droplets. This osmotic force increases with
or hydrogen bond, between WPI and KGM. However, the droplet increasing concentration of polymer until it is large enough to
size (data not shown) of WPI-KGM dispersions significantly overcome the repulsive forces between droplets and cause floccu-
decreased after homogenization, which was mainly attributed to lation of droplets. Droplet flocculation can also increase the vis-
the mechanical de-polymerization of KGM by homogenization, as cosity of emulsions by decreasing the internal packing of droplets
described above (Villay et al., 2012). within flocs due to increased effective volume fraction of the par-
ticles based on Dougherty-Krieger equation (McClements, 2015).
This may explain why emulsion with high content of KGM showed
3.4. Creaming stability
increased viscosity, as described above (Fig. 1).
Creaming index (CI) was used to describe the rate of light
transmission change. It is calculated based on the curve of the in- 3.5. Stability of emulsions at different pH
tegrated transmitted light against time. A higher CI value indicates
a lower creaming stability of emulsion. Hence, emulsion containing Emulsions were all stable at pH 2e4 and pH 6e7. The emulsion
0.05% KGM showed the highest creaming stability, followed by without KGM showed a significantly increase in droplet size at pH 5
emulsions containing 0.1% KGM and the emulsion without KGM (Fig. 4a). This is mainly attributed to the aggregation of whey
(Table 1). KGM can improve the creaming stability of emulsions by proteins at this pH value, which is close to its isoelectric point (IEP).
increasing the viscosity of the water phase as described above This increase in droplet size accordingly resulted in poor creaming
(Fig. 1). In addition, KGM at these concentrations (0.05%, or 0.1%) stability of these emulsions (Table 2). Emulsions containing KGM
can potentially form a network structure in water phase. Even showed significantly less increase in droplet size at pH 5 (p < 0.05)
though, these network structures can be destroyed by mechanical than the emulsion without KGM, depending on the content of KGM.
homogenization as described above, the chain-structure of KGM Accordingly, these KGM-containing emulsions with reduced
can still limit the creaming of emulsions due to the increased steric droplet size at pH 5 showed better creaming stability than the
hindrance and force of friction between droplets and the contin- emulsion without KGM (Table 2). These results suggested that KGM
uous phase. can significantly improve the pH stability of WPI-stabilized
However, emulsion containing 0.2% KGM showed very poor emulsions.

Table 2
Creaming index (CI) of emulsions at pH 5.0 or after freeze-thaw, oil-off of emulsions after freeze-thaw, and droplet size of emulsions at different temperatures.

KGM content CI at pH 5.0 CI after freeze-thaw (%/min) Oil-off after Size (d.nm)
(%, w/w) (%/min) freeze-thaw (%)
25
C 37
C 80
C

0 9.11 ± 0.05a 0.76 ± 0.00b 10.2 ± 0.03a 252±8d 254±5d 257±2d

0.05 7.77 ± 0.03b 0.70 ± 0.00c 6.7 ± 0.02b 271 ± 10c 269±9c 266±6c
0.1 5.98 ± 0.03c 0.62 ± 0.00d 4.8 ± 0.00c 313±6b 312±6b 314±6b
0.2 3.88 ± 0.04d 1.36 ± 0.01a 2.4 ± 0.02d 334±6a 328±6a 331±6a

*KGM indicates konjac glucomannan.

a
Different letters indicate significant difference between values in a column (p < 0.05).

Table 3
Droplet size and surface charge of emulsions after freeze-thaw processing.

Freeze-thaw Size (d.nm) Zeta potential (mV)

cycle
KGM 0% KGM 0.05% KGM 0.1% KGM 0.2% KGM 0% KGM 0.05% KGM 0.1% KGM 0.2%

0 252±9a 280±8a 306 ± 17a 350 ± 24a 65.0 ± 5.4a 60.4 ± 2.2a 59.3 ± 2.6a 59.3 ± 0.7a
1 202±4b 226±2b 266 ± 18b 302 ± 18b 57.3 ± 0.1b 52.6 ± 1.4b 53.1 ± 2.1b 52.8 ± 0.7b
2 216±5c 249±8c 288 ± 16c 308 ± 13b 52.5 ± 1.8c 54.5 ± 1.8b 53.9 ± 1.8b 52.4 ± 1.3b
3 227±6d 273±9d 307±9d 314 ± 10b 52.5 ± 1.7c 54.2 ± 1.4b 51.4 ± 0.1b 54.5±0b

* Freeze-thaw test was done at 20
C, followed by thawing at 25
C. KGM 0% indicate emulsion without konjac glucomannan (KGM). KGM 0.05%, KGM 0.1%, and KGM 0.2%
indicates emulsions with KGM content (w/w) of 0.05%, 0.1%, and 0.2%, respectively. aDifferent letters indicate significant difference between values in a column (p < 0.05).

Table 4
Droplet size and zeta potential of emulsions containing KGM following different phases of gastrointestinal tract digestion.

KGM content (w/w, %) Size (d.nm) zeta potential (mV)

initial mouth phase gastric phase intestinal phase initial mouth phase gastric phase intestinal phase

0 252±9d 248±1d 1505 ± 361a 192 ± 56a 65 ± 5.4a 64.5 ± 4.9a 18.0 ± 1.4a 70.5 ± 6.4a
0.05 280±8c 288±1c 1624 ± 452a 185 ± 31a 60.4 ± 2.2a 58.5 ± 0.7a 15.5 ± 0.7a,b 71.0 ± 2.8a
0.1 306 ± 17b 305±4b 1598 ± 423a 178 ± 19a 59.3 ± 2.6a 59.5 ± 2.1a 14.0±0b 76.5 ± 4.9a
0.2 350 ± 24a 352 ± 11a 1057 ± 204a 200 ± 38a 59.3 ± 0.7a 59.0 ± 1.4a 9.0±0c 77.5 ± 3.5a

*KGM indicates konjac glucomannan.

a
Different letters indicate significant difference between values in a column (p < 0.05).
126 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128

KGM is a very stable polysaccharide across a wide range of pH charge of all emulsions significantly decreased (p < 0.05) (Table 3).
values. KGM, with a chain-like structure, can fill in the space be- Then, the droplet size increased with the increasing freeze-thaw
tween free protein molecules and oil droplets in emulsions and cycles and the final droplet size of all emulsions after three cycles
potentially can isolate oil droplets from each other, which accord- were smaller than their initial droplet size. No significant difference
ingly reduces the Brownian-motion-induced contact of proteins in surface charge after second and third cycles of freeze-thaw was
and oil droplets and thus their aggregation at pH value close to the observed, except for the emulsion without KGM. Oiling-off of
IEP of proteins. The higher of the KGM content, the higher density of emulsions was also clearly observed after 3 cycles of freeze-thaw
the filled space, and thus the more significant inhibition of (Table 2) and KGM successfully reduced the oiling-off of the
aggregation. emulsion. Compared with the emulsion without KGM (10.2%), only
2.4% of the oil was released from the emulsion containing 0.2% KGM
after 3 cycles of freeze-thaw (p < 0.01).
3.6. Thermal stability
Ice crystals are formed during freezing of emulsions, and the ice
penetration can potentially lead to the break-down of the protein
The thermal stability of emulsions at different temperatures was
layer surrounding the oil droplets, and thus lead to oiling-off of
also investigated. After 2 h incubation at elevated temperature
emulsions. Combined with the results above, it is assumed that (i)
(37
C or 80
C), the droplet size of all emulsions showed almost no
freezing led to the break-down of large emulsion droplets, leading
difference compared to the droplet size of emulsions at 25
C
to the oiling-off of emulsions, a narrower size distribution and a
(Table 2), suggesting that emulsions in this study had good thermal
smaller average droplet size, and (ii) freezing can also induce
stability at temperatures up to 80
C.
desorption of whey proteins from the droplet surface, leading to the
reduced surface charge of emulsions after first cycle of freeze-thaw.
3.7. Freeze-thaw testing However, the above findings suggested that KGM can help to
maintain the structure of the interfacial protein layer and
After the first cycle of freeze-thaw, the droplet size and surface

Fig. 5. Confocal scanning microscope observation of emulsion droplets after being exposed to simulated gastrointestinal tract (GIT). KGM 0% indicate emulsion without konjac
glucomannan (KGM). KGM 0.05%, KGM 0.1%, and KGM 0.2% indicates emulsions with KGM content (w/w) of 0.05%, 0.1%, and 0.2%, respectively.
 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128 127

significantly protect the emulsion from oiling-off, probably by influence the four steps of the digestion process of emulsion
slowing down the formation of ice crystals (Mao et al., 2015). droplets can potentially modify the final release of encapsulated
ingredients from emulsions. Our previous studies also showed that
emulsions with maltodextrin-structured water phase showed
3.8. Simulated GIT digestion
significantly different release profiles of lipophilic food flavors due
to modified mobility within the water phase (Mao, Roos, & Miao,
As described in our previous studies (Lu, Kelly, & Miao, 2017a; Lu
2014). In this study, KGM was used to formulate the water phase
et al., 2017b), passing emulsions through simulated gastrointestinal
of model O/W emulsion. Emulsions with KGM showed more a
tract (GIT) digestion can greatly modify their properties. The
viscous water phase than emulsion without KGM. In addition, KGM
average droplet size of all emulsions significantly increased after
shows a chain-like molecular structure in water, which facilitates
gastric phase and then decreased after intestinal phase (Table 4),
intermolecular cross-linking into a gel-like structure. All these
which was also confirmed by the microscopy observation. As is
properties can potentially interfere with the digestion steps of the
shown in Fig. 5, after the mouth phase, no major difference in
hydrolysis of interfacial protein layer and the hydrolysis of oil phase
droplet shape and size was observed. After the gastric phase, sig-
by steric hindrance effect, leading to slower release of encapsulated
nificant aggregation of proteins (green fluorescence) was observed,
ingredients from emulsion droplets. These factors may explain why
which was mainly attributed to the exposure of the hydrophobic
emulsions with KGM-structured water phase showed decreased
domain of whey proteins induced by pepsin hydrolysis and rela-
release rates of encapsulated b-carotene with increasing KGM
tively high ionic strength in gastric phase. The process of protein
content in this study. The findings also suggest that it is feasible to
aggregation also clustered some oil droplets (red fluorescence) and
achieve a controlled/sustainable release of encapsulated functional
formed larger complexes. In addition, hydrolysis of WPI at the
ingredients from emulsions by structuring the water phase of
interface by pepsin resulted in aggregation of some oil droplets.
emulsions with natural biopolymers.
These two factors may explain the dramatic increase in droplet size
However, KGM can be hydrolysed and utilized by the microor-
after the gastric phase. After the intestinal phase, almost all pro-
ganisms in the gut. Hence, the in-vivo digestion behavior and
teins at the interface were hydrolysed by trypsin, leading to the
release profile of encapsulated ingredients in emulsions with a
collapse of oil droplets and the hydrolysis of oil by lipase. The hy-
KGM-structured water phase need to be further studied.
drolysis of whey protein and sunflower oil also led to the quenching
of most of the green and red fluorescence, respectively. Only very
few intact oil droplets were captured. 4. Conclusion

Konjac glucomannan (KGM) increased the viscosity of the water

3.9. Release of encapsulated b-carotene after GIT
phase and thus the viscosity of the final whey-protein-isolate (WPI)
stabilized O/W emulsions. No significant non-covalent interaction
Adding KGM to the water phase of emulsions significantly
between KGM and whey proteins before and after homogenization
modified the release of b-carotene from emulsion droplets after
was observed by FT-IR. KGM greatly improved the creaming and pH
passing through GIT. Emulsion without KGM showed a final b-
stability of emulsions and protected emulsions from oiling-off after
carotene-release-rate of 64.5%. b-carotene-release-rate of emul-
freeze-thaw process. The digestion and break-down of emulsion
sions containing KGM were generally inferior to that of emulsion
droplets mainly happened in the intestinal phase of a simulated
without KGM, and the release rate decreased with increasing KGM
gastrointestinal tract (GIT), as evaluated by confocal laser scanning
content (Fig. 6). Emulsions containing 0.1% KGM and 0.2% KGM
microscopy. KGM significantly decreased the release rate ofb-
showed significant lower b-carotene-release-rate than emulsion
carotene from emulsions, which is dependent on the content of
without KGM (p < 0.05). The results indicated that KGM can
KGM in emulsions.
potentially slow down the release of b-carotene from emulsion
Model O/W emulsions with better stability and controlled
droplets, which is dependent on the content of KGM in final
release of b-carotene were obtained by simply structuring the
emulsions.
water phase with a health-beneficial polysaccharide, KGM. Find-
As discussed in the introduction section, any factors that
ings in this study make it possible to design emulsion-based
functional food products or drug carriers with potential
controlled or sustainable release of functional ingredients inside by
structuring the water phase of emulsions with natural edible
biopolymers.

Funding

This work was supported by National Natural Science Founda-

tion of China (No.31628016), China Scholarship Council
(No.201508300001), and Teagasc-The Irish Agriculture and Food
Development Authority (RMIS6821)

Notes

The authors declare no conflict of interest.

Acknowledgement

Fig. 6. Release of encapsulated b-carotene from emulsions. KGM indicates konjac The assistance of Ana Isabel Mulet Cabero with the laser scan-
glucomannan. *p < 0.05, **p < 0.01. ning confocal microscope observation was highly acknowledged.
128 W. Lu et al. / Food Hydrocolloids 81 (2018) 120e128

References vegetable proteins with other polymers: Structure-function relationships and

applications in the food industry. Trends in Food Science & Technology, 68,
130e144.
Aoki, T., Decker, E. A., & McClements, D. J. (2005). Influence of environmental
Lu, W., Kelly, A. L., Maguire, P., Zhang, H., Stanton, C., & Miao, S. (2016). Correlation of
stresses on stability of O/W emulsions containing droplets stabilized by
emulsion structure with cellular uptake behavior of encapsulated bioactive
multilayered membranes produced by a layer-by-layer electrostatic deposition
Nutrients: Influence of droplet size and interfacial structure. Journal of Agri-
technique. Food Hydrocolloids, 19(2), 209e220.
cultural and Food Chemistry, 64, 8659e8666.
Arvill, A. B., & Lennart. (1995). Effect of short-term ingestion of konjac glucomannan
Lu, W., Kelly, A. L., & Miao, S. (2016). Emulsion-based encapsulation and delivery
on serum cholesterol in healthy men. American Journal of Clinical Nutrition, 61,
systems for polyphenols. Trends in Food Science & Technology, 47, 1e9.
585e589.
Lu, W., Kelly, A. L., & Miao, S. (2017a). Bioaccessibility and cellular uptake of beta-
Barth, A. (2007). Infrared spectroscopy of proteins. Biochimica et Biophysica Acta,
carotene encapsulated in model O/W Emulsions: Influence of initial droplet
1767(9), 1073e1101.
size and emulsifiers. Nanomaterials, 7(9), 282e292.
Chen, B., Li, H., Ding, Y., & Suo, H. (2012). Formation and microstructural charac-
Lu, W., Kelly, A. L., & Miao, S. (2017b). Improved bioavailability of encapsulated
terization of whey protein isolate/beet pectin coacervations by laccase cata-
bioactive nutrients delivered through monoglyceride-structured O/W emul-
lyzed cross-linking. Lebensmittel-Wissenschaft und -Technologie- Food Science
sions. Journal of Agricultural and Food Chemistry, 65, 3048e3055.
and Technology, 47(1), 31e38.
Mao, Y., E. C, M., & Lekkerkerker, H. N. W. (1995). Depletion force in colloidal sys-
Chua, M., Baldwin, T. C., Hocking, T. J., & Chan, K. (2010). Traditional uses and po-
tems. Physica a, 222, 10e24.
tential health benefits of Amorphophallus konjac K. Koch ex N.E.Br. Journal of
Mao, L., Roos, Y. H., & Miao, S. (2014). Study on the rheological properties and
Ethnopharmacology, 128(2), 268e278.
volatile release of cold-set emulsion-filled protein gels. Journal of Agricultural
Dickinson, E. (2011). Mixed biopolymers at interfaces: Competitive adsorption and
and Food Chemistry, 62(47), 11420e11428.
multilayer structures. Food Hydrocolloids, 25(8), 1966e1983.
Mao, L., Roos, Y. H., & Miao, S. (2015). Effect of maltodextrins on the stability and
Gao, Z., Su, R., Huang, R., W. Q, & He, Z. (2014). Glucomannan-mediated facile
release of volatile compounds of oil-in-water emulsions subjected to
synthesis of gold nanoparticles for catalytic reduction of 4-nitrophenol. Nano-
freezeethaw treatment. Food Hydrocolloids, 50, 219e227.
scale Research Letters, 9, 8.
McClements, D. J. (2015). Food emulsions: Principles, practices, and techniques (3rd
Guzey, D., Kim, H. J., & McClements, D. J. (2004). Factors influencing the production
ed.). New York: CRC press.
of o/w emulsions stabilized by b-lactoglobulinepectin membranes. Food Hy-
Niu, Z., Conejos-Sanchez, I., Griffin, B. T., O'Driscoll, C. M., & Alonso, M. J. (2016).
drocolloids, 18(6), 967e975.
Lipid-based nanocarriers for oral peptide delivery. Advanced Drug Delivery Re-
Jeonghee Surh, Y. S. G., Decker, Eric A., & Julian McClements, D. (2005). Influence of
views, 106(Pt B), 337e354.
environmental stresses on stability of emulsions containing cationic droplets
Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2012). Nanoemulsion delivery
stabilized by SDS fish Gelatin membranes. Journal of Agricultural and Food
systems: Influence of Carrier oil on beta-carotene bioaccessibility. Food Chem-
Chemistry, 53, 4236e4244.
istry, 135(3), 1440e1447.
Karthik, P., & Anandharamakrishnan, C. (2016). Enhancing omega-3 fatty acids
Salvia-Trujillo, L., Fumiaki, B., Park, Y., & McClements, D. J. (2017). The influence of
nanoemulsion stability and in-vitro digestibility through emulsifiers. Journal of
lipid droplet size on the oral bioavailability of vitamin D2 encapsulated in
Food Engineering, 187, 92e105.
emulsions: An in vitro and in vivo study. Food Funct, 8, 767e777.
Klinkesorn, U., Sophanodora, P., Chinachoti, P., & McClements, D. J. (2004). Stability
Villay, A., Lakkis de Filippis, F., Picton, L., Le Cerf, D., Vial, C., & Michaud, P. (2012).
and rheology of corn oil-in-water emulsions containing maltodextrin. Food
Comparison of polysaccharide degradations by dynamic high-pressure ho-
Research International, 37(9), 851e859.
mogenization. Food Hydrocolloids, 27(2), 278e286.
Lazaridou, A., Biliaderis, C. G., & Izydorczyk, M. S. (2003). Molecular size effects on
Ye, A., Hemar, Y., & Singh, H. (2004). Enhancement of coalescence by xanthan
rheological properties of oat b-glucans in solution and gels. Food Hydrocolloids,
addition to oil-in-water emulsions formed with extensively hydrolysed whey
17(5), 693e712.
proteins. Food Hydrocolloids, 18(5), 737e746.
Lin, D., Lu, W., Kelly, A. L., Zhang, L., Zheng, B., & Miao, S. (2017). Interactions of