-
ORIGINAL PAPER
Evaluating the Effectiveness of β-Carotene Extractionfrom Pulsed
Electric Field-Treated Carrot Pomace UsingOil-in-Water
Microemulsion
S. Roohinejad & I. Oey & D. W. Everett & B. E.
Niven
Received: 2 January 2014 /Accepted: 8 May 2014 /Published
online: 18 May 2014# Springer Science+Business Media New York
2014
Abstract Thermodynamically stable microemulsions wereused to
extract β-carotene from pulsed electric field (PEF)-treated carrot
pomace. In this study, a three-level Box–Behnken design was used to
predict the effect of extractiontime (10–110 min), extraction
temperature (30–70 °C) andcarrot/microemulsion ratio (1:30–1:90w/w)
on the β-carotenecontent, polydispersity index (PDI) and particle
size of themicroemulsions. The β-carotene extracted from
PEF-treatedcarrot pomace using microemulsions was higher than
untreat-ed carrot pomace. The extraction efficiency of
β-caroteneusing microemulsions was higher compared to 100 %
hexaneor 100 % glycerol monocaprylocaprate oil. A mathematicalmodel
was developed to predict the optimal extraction condi-tions using
transparent microemulsions with high loading ofβ-carotene, low PDI
and small microemulsion particle size.The model predicted that an
extraction time of 49.4 min,temperature of 52.2 °C and
carrot/microemulsion ratio of1:70 (w/w) would result in
microemulsions with β-caroteneloading of 19.6 μg/g, PDI of 0.27 and
particle size of 74 nm.This study demonstrates the potential of
using oil-in-watermicroemulsions as extraction media for
β-carotene.
Keywords Microemulsion . Extraction . Polydispersityindex .
Particle size . Optimization .β-Carotene
Introduction
Interest in the replacement of synthetic pigments bynatural
colourants extracted from plant materials formaking transparent
beverages has been and is increasingin recent years due to high
consumer demand for ‘morenatural’ and healthy beverages (Cai et al.
2005; Lópezet al. 2009). β-Carotene is the most abundant
caroten-oid, which is found in high concentrations in manyfresh
fruits and vegetables, such as carrots. The impor-tance of
β-carotene in food goes beyond their role asnatural colourants;
there are well-documented biologicalactivities such as
strengthening the immune system,decreasing the risk of cancer and
reducing the risk ofcoronary heart diseases (Van Poppel and
Goldbohm1995; Kritchevsky 1999).
The physical state and location of carotenoids inplants strongly
affects their accessibility during diges-tion, which subsequently
limits release and absorption.This can occur due to deposition of
carotenoids incrystalline form in the chromoplasts, or carotene
crys-tals being stabilized by other components (Reiter et al.2003).
Various techniques can be used to break downplant cell structure
and improve the release of caroten-oids, such as mechanical
homogenization and heattreatment (Hedrén et al. 2002); however,
these tech-niques may promote reactions that cause the degrada-tion
of carotenoids (Fratianni et al. 2010; Talcott andHoward 1999).
Organic solvents are normally used to extract
oil-solublecompounds from cellular structures after employing
physicalmethods to improve extractability. Conventional
carotenoidextraction methods require large amounts of organic
solvents,which are costly, environmentally hazardous, and
requireexpensive disposal procedures (Mustafa et al.
2012).Furthermore, the traditional extraction of carotenoids is
no
S. Roohinejad : I. Oey (*) :D. W. EverettDepartment of Food
Science, University of Otago, PO Box 56,Dunedin 9054, New
Zealande-mail: [email protected]
D. W. EverettRiddet Institute, Private Bag 11 222, Palmerston
North 4442,New Zealand
B. E. NivenCentre for Application of Statistics and Mathematics,
University ofOtago, PO Box 56, Dunedin 9054, New Zealand
Food Bioprocess Technol (2014) 7:3336–3348DOI
10.1007/s11947-014-1334-6
-
longer recommended because of the risk of organic
solventresidues and loss of carotenoids as a consequence of
solventevaporation (Illés et al. 1999). Recently, alternative
non-thermal processing methods such as pulsed electric field(PEF)
have emerged to improve the release of carotenoidsthrough
increasing the permeabilization of cells (Grimi et al.2007;
Roohinejad et al. 2014) without any significant effecton the
stability of β-carotene or total carotenoid content(Sanchez-Moreno
et al. 2005). Moreover, the use of inexpen-sive and efficient
alternate extraction systems, such asmicroemulsions, might be
suitable to replace the use of or-ganic solvents.
Microemulsions are defined as a system formed bythe dispersion
of microdroplets of two immiscible liq-uids, stabilized by an
interfacial membrane formed bysurfactants. They are
thermodynamically stable, homo-geneous and optically isotropic
solutions with an aver-age droplet size less than 100 nm. These
delivery sys-tems are of great technological and scientific
interestbecause of their potential to incorporate a wide rangeof
bioactives (hydrophilic and hydrophobic) due to thepresence of both
lipophilic and hydrophilic domains.These adaptable delivery systems
provide protectionagainst oxidation, enzymatic hydrolysis and
improvethe solubilization of lipophilic bioactives, hence
enhancetheir bioavailability (Talegaonkar et al. 2008; Flanaganand
Singh 2006).
Previous studies reported on the ability of microemulsionsto
protect carotenoids against oxidation and increase solubili-zation
(Arvanitoyannis 2009; Spernath et al. 2002).
Moreover,microemulsions have attracted attention because of their
ca-pability for selective extraction of biomolecules and metalions
in liquid–liquid systems (Cortez et al. 2004; Dantaset al. 2003)
and DNA condensation (Budker et al. 2002);however, there are very
few reports on the potential ofmicroemulsions to extract food
components from food com-plex mixtures (Paul and Moulik 2001). The
utilization ofmicroemulsified systems for β-carotene extraction
might bea preferable alternative to conventional solvent extraction
dueto less hazardous solvents and lower energy consumption.
This study was designed to (1) evaluate the effectiveness
ofoil-in-water (o/w) microemulsions as a potential alternative
toreplace the use of organic solvents for extracting β-carotenefrom
carrot pomace and (2) optimize the process conditions(extraction
time, temperature and ratio) on dependent re-sponses (β-carotene
content, polydispersity index (PDI) andparticle size) using a
Box–Behnken design. The effect of PEFtreatment on the
extractability of β-carotene from carrot pom-ace using
microemulsions was initially compared to non-PEF-treated carrot
pomace. The optimized extraction conditions toobtain the maximum
β-carotene content, minimum PDI andsmall particle size were
determined. To our knowledge, this isthe first study that has used
microemulsions as an
environmentally friendly mixture for extraction of β-carotene
from plant materials for use in making transparentbeverages.
Materials and Methods
Reagents
Methyl t-butyl ether and Tween 80 (polyoxyethylene
sorbitanmonooleate) were purchased from Sigma-Aldrich (St.
Louis,MO, USA) and n-hexane from J.T. Baker (Phillipsburg, NJ,USA).
Glycerol monocaprylocaprate (Capmul MCM; CASno. 26402-26-6) was
kindly donated by Abitec Corporation(Janesville, WI, USA). The
fatty acid composition of CapmulMCM consists of caprylic acid (C8),
capric acid (C10) andlauric acid (C12) in the ratio of
97.3:2.6:0.1. Ethanol (95 %),methanol (HPLC-grade) and acetone were
purchased fromBiolab (Scoresby, Victoria, Australia). A β-carotene
standardfor HPLC analysis was purchased from
CaroteNature(Lupsingen, Switzerland). All solvents used in this
experimentwere HPLC-grade, with the exception of ethanol.
Preparation of Pulsed Electric Field-Treated Carrot Pomace
Fresh carrots (Daucus carota cv. Nantes) harvested betweenJune
and July 2013 were purchased from a supplier of locallygrown
products (Kaan's Catering Supplies Ltd., Dunedin,New Zealand).
Samples were washed, sliced and processedinto a purée by mixing
with distilled water (1:1 ratio). PEFtreatment was carried out and
compared with untreated sam-ples (control) of the corresponding
batches. Carrot purée wastreated using an ELRACK-HVP 5 PEF unit
(Quakenbrück,Germany) in a batch treatment configuration. For each
PEFtreatment, carrot purée (100 g) was placed inside the PEFchamber
of dimensions 80×100×50 mm with a sample ca-pacity of 400 mL,
consisting of two stainless steel parallelplate electrodes with a
gap of 80 mm. The treatment wasapplied according to the following
optimum conditions: elec-tric field strength of 0.6 kV/cm with
constant frequency of5 Hz, pulse width of 20 μs and treatment time
of 3 ms, asdetermined from a previous study (Roohinejad et al.
2014).The temperature of the carrot purée was measured prior to
andafter PEF treatment, using an electric thermometer
(BAT-10thermometer, Physitemp, Clifton, NJ, USA). All samples
weretreated using bipolar square pulses. PEF-treated carrot
puréewas centrifuged twice at 15,300×g for 30 min at 4 °C, and
thesolid phase (pomace) was separated and freeze-dried. Afterthis
process, the dehydrated carrot pomace (carrot fibre) wasmilled,
passed through a 14 mesh (1.19 mm) sieve and storedat −18 °C for
later use. Untreated carrot pomace was alsofreeze-dried, milled,
sieved and stored as for the PEF-treatedcarrot pomace.
Food Bioprocess Technol (2014) 7:3336–3348 3337
-
Experimental Design
A preliminary study was carried out to determine the
optimalβ-carotene extraction efficiency from untreated carrot
pomace(UCP) and PEF-treated carrot pomace (PCP) using an
o/wmicroemulsion (Capmul MCM/Tween 80/phosphate buffer(30:20:50,
w/w)). The optimum extraction conditions weredetermined by means of
a three-factor, three-level Box–Behnken experimental design
combined with response sur-face modelling (RSM) and quadratic
programming. RSMconsists of a group of empirical techniques which
are appro-priate for the evaluation of relationships between
controlledexperimental factors andmeasured responses according to
oneor more selected criteria (Bayraktar 2001). Using a Box–Behnken
design instead of a complete factorial design hasmany advantages,
including (i) they are all spherical designsand factors with only
three levels required to be run, (ii) someof these designs can
provide orthogonal blocking, (iii) thereare no runs where all
factors are at either the maximum orminimum levels and (iv)
additional experiments and timeconsuming laboratory work can be
eliminated (Yetilmezsoyet al. 2009).
The conditions for optimized extraction were time of ex-traction
(X1), temperature of extraction (X2) and carrot pomaceto
microemulsion ratio (X3), at three equidistant levels. Theresponse
variables were the β-carotene content (Y1), PDI (Y2)and particle
size (Y3). A regression model containing tencoefficients including
linear and quadratic effect of factorsand linear effect of
interactions was assumed to describerelationships between
experimental factors (X1, X2 and X3)and the responses (Y1, Y2 and
Y3) as follows:
Yk ¼ β0 þX
i¼1
3
βiX i þX
i¼1
3
βiiX2i þ
X
i¼1
2 X
j¼iþ1
3
βijX i:X j
ð1Þ
where β0, βi, βii and βij are the intercept, regression
coeffi-cients of the linear, quadratic and interaction terms of
themodel, respectively, and Xi and Xj are the independent
vari-ables and Y is the dependent variable (β-carotene content,
PDIand particle size). The combination of factors at the centre
oflevel was run in triplicate. The numbers of experiments (N)were
calculated as follows:
N ¼ 2 k k−1ð Þ þ cp ð2Þ
where k is the number of factors and cp is the number
ofreplicates at the central point. In total, 15 combinations
ofthree factors were used. The extraction of carotenoids
anddetermination of β-carotene was carried out using HPLC.
All experiments were carried out in a randomized order
toeliminate bias. The factors, their levels and codes for the
levelsare shown in Table 1. The 3D response graph and profile
forpredicted values and desirability level for factors were
plottedusing Minitab software (version 16, Minitab Inc., State
Col-lege, PA, USA).
β-Carotene Extraction Using a Microemulsion
Pseudo-ternary phase diagrams containing oil–surfactant–wa-ter
mixtures were constructed to define the microemulsionregion.
Briefly, samples were prepared by mixing appropriateamounts of
medium chain triglycerides (Capmul MCM),phosphate buffer (0.01 M,
pH 6.8) and surfactant (Tween80) in vials at room temperature and
mixed well using a vortexmixer. Buffers were prepared using
NaH2PO4/Na2HPO4.NaOH (0.1 M) or HCl (0.1 M) were used to maintain
the pHof the solution. The pseudo-ternary phase diagrams
wereconstructed by preparing 100 samples of differing composi-tions
to define the phase boundaries in each phase diagram,using
SigmaPlot software (Version 12.3, Systat Software Inc.,Chicago, IL,
USA). A clear, transparent o/w microemulsionliquid region with a
maximum oil loading and minimumparticle size (30 % Capmul MCM and
20 % Tween 80 in50 % 0.01 M, pH 6.8 phosphate buffer) was selected
for thisexperiment. Freeze-dried carrot pomace (UCP and PCP)
weremixed with microemulsions, flushed with nitrogen for 1
min,placed in a thermostatic orbital shaker (Model OM 11,
RatekInstrument Ltd., Boronia, Victoria, Australia) and shaken
at300 rpm continuously at different temperatures (30, 40, 50, 60and
70 °C), times (10, 20, 30, 60, 90, 110 min), and carrotpomace to
microemulsion ratios (1:30, 1:40, 1:60, 1:70 and1:90). Freeze-dried
carrot pomace was used instead of carrotpomace due to the reason
that water present in the carrotpomace will change the percentage
of water in the formulationwhich consequently affects microemulsion
transparency.After completing the extraction process, the
suspensions were
Table 1 Levels of independent and dependent variables in
Box-Behnkendesign
Independent variables Level
Low (−1) Middle (0) High (+1)
X1=time of extraction (min) 10 60 110
X2=temperature of extraction (°C) 30 50 70
X3=PCP/microemulsion ratio (w/w) 1:30 1:60 1:90
Dependent variables Goal
Y1=β-carotene Maximize
Y2=PDI Minimize
Y3=particle size Minimize
PCP pulsed electric field-treated carrot pomace, PDI
polydispersity index
3338 Food Bioprocess Technol (2014) 7:3336–3348
-
transferred to 50 mL Polyallomer centrifuge bottle
(Beckman,Fullerton, CA, USA) and centrifuged at 12,100×g for 15
minat 4 °C and the clear supernatant liquid was decanted. The
β-carotene content, PDI and droplet size were determined usingthe
methods described below.
Microemulsion Droplet Size, PDI and Zeta Potential Analysis
The particle size and PDI of the blank and
carotenoid-loadedmicroemulsions were determined without any
dilution by adynamic light scattering technique. Light scattering
measure-ments were carried out using a high-performance particle
sizer(HPPS; model HPP5001, Malvern Instruments Ltd.,Worcestershire,
UK). The mean particle size was reported asa Z-average, which is
the mean hydrodynamic diameter of theparticle derived from the
cumulative analysis of the intensityof the scattered light
(Schmidts et al. 2011). The zeta potentialof the microemulsions was
determined using a zeta potentialanalyser (Zetasizer 3000, Malvern
Instruments Ltd.) (Fanet al. 2014).
Microemulsion Stability
The stability of the optimized microemulsion prepared by
theBox–Behnken mathematical model used in this study wasevaluated
by analysing the β-carotene content, droplet sizeand PDI. After
preparation, this microemulsion was stored at25 °C for 4 weeks and
its stability was evaluated after storagefor 0, 2 and 4 weeks and
compared with the predicted valuegiven by the model. If the
microemulsions are not stable,separation or turbidity will be
observed to increase duringthe storage time.
Encapsulation Efficiency
The β-carotene encapsulation efficiency of
optimizedmicroemulsion was measured according to a method
de-scribed by Qu et al. (2014). The microemulsion was pouredonto a
0.22 μm cellulose nitrate membrane (Sartorius,Göttingen, Germany)
to remove the nonencapsulated β-carotene. The free β-carotene was
retained in the membranewhile the filtrate containing the
homogeneous suspension ofβ-carotene-loaded microemulsion was
collected. The β-carotene encapsulation efficiency (DEE) was
calculated asfollows:
DEE ¼ C1C2
� 100 ð3Þ
whereC1 represents the measured amount ofβ-carotene in
themicroemulsion after filtration, and C2 is the total amount
ofβ-carotene in the microemulsion before filtration.
Determination of Carotenoid Content in the Microemulsion
The extraction of carotenoids using a microemulsion wascarried
out by solvent extraction (Wright et al. 2008).Briefly, samples
were subjected to solvent extraction byadding 0.5, 3 and 1 mL of
ethanol, acetone and deionizedwater, respectively, with 5 s of
vortexing after the addition ofeach liquid. Hexane (2 mL) was
added, the vials inverted tentimes and the organic layer was
removed after 5 min using aglass transfer pipette. The hexane
extraction was carried out intriplicate for each sample. The hexane
extraction (6 mL) waspooled and the organic phase was dried under
nitrogen at37 °C to the point of dryness. HPLC-grade methanol (1
mL)was added and vortexed for 5 s. The β-carotene content
wasanalysed by HPLC after extraction.
The HPLC assay for the identification and quantification
ofβ-carotene was based on the method of Lemmens et al.(2009). The
separation of the β-carotene was carried out witha reversed phase
C18-column (5 μm×250 mm×4.6 mm). Forthis purpose, an HPLC system
equipped with a diode arraydetector (Agilent Technologies 1200
Series, Diegem,Belgium) was used. A linear gradient elution was
applied toseparate the carotenoids. The gradient, starting from the
initialcondition (81 % methanol, 15 % methyl t-butyl ether, 4
%reagent grade water) was built up in 20 min to the endcondition
(41 % methanol, 55 % methyl t-butyl ether, 4 %reagent-grade water)
at a flow rate of 1 mL/min. Identificationand quantification was
carried out at 450 nm, at which pointthe maximal absorbance of
β-carotene was found. The col-umn temperature and the autosampler
temperature were set at25 and 4 °C, respectively. The mean value of
β-carotene wasobtained from at least three independent extractions
using aβ-carotene external standard curve.
Transmission Electron Microscopy
The morphology of the optimized microemulsions (blank
andβ-carotene-loaded microemulsion) was examined using
trans-mission electron microscopy (TEM). Samples (10 μL) wereplaced
on plasma-glowed, carbon-coated, 300 mesh coppergrids (ProSciTech
Pty Ltd, Thuringowa, Queensland,Australia). The carbon coating and
plasma glowing steps werecarried out using an Edwards E306A coating
system(Edwards Vacuum Ltd, Crawley, England). Plasma glowingrenders
the carbon film hydrophilic immediately before spec-imen
application to create an optimal specimen film. After60 s, the
excess specimen was blotted from the grid. Thesample residue was
contrasted using 1 % phosphotungsticacid at pH 6.8 before being
immediately blotted again.Images were captured using a Philips
CM100 BioTWINtransmission electron microscope (Philips/FEI
Corporation,Eindhoven, The Netherlands) with a Mega View III
digitalcamera (Olympus Soft Imaging Solutions GmbH, Münster,
Food Bioprocess Technol (2014) 7:3336–3348 3339
-
Germany). The TEM was operated at 100 kV acceleratingvoltage,
and typically viewed at a magnification of×13,500.
Statistical Analysis
Three replications of independent experiment were carried outand
triplicate samples of each experiment were analysed. Atwo-sample t
test was used to compare the β-carotene con-centrations of the
microemulsions obtained from untreatedand PEF-treated carrot
pomace. The significance of indepen-dent variables and their
interactions were analysed usingANOVA. The adequacy of the
quadratic model was verifiedby ANOVA, determination coefficient
(R2), adjusted determi-nation coefficient (Ra
2), correlation coefficient (R) and chi-square (χ2) tests. All
statistical analyses were carried out usingMinitab software
(version 16, Minitab Inc., State College, PA,USA).
Results and Discussion
Evaluation of Method
Effect of Pulsed Electric Field on Release of β-Carotene
To compare the extraction efficiency of carotenoids from PCPand
UCP, a direct carotenoid extraction was carried out. Atwo-sample t
test was used to compare with the extraction ofβ-carotene from PCP
and UCP. In comparison to the untreatedsample, PEF treatment led to
an insignificant (P>0.05) in-crease in temperature when the
final temperature was below25 °C. The results showed that the
β-carotene extracted fromPCPwas significantly (P
-
100% hexane and Capmul oil (the oil phase component of
themicroemulsion). The microemulsion used in this study
onlycontained 30 % oil and was able to maximally extract (up
to12.75 %)β-carotene at 22.80±0.63μg/g for UCP and 24.79±0.62 μg/g
for PCP (Fig. 1). The β-carotene content of PCPand UCP was also
measured after hexane extraction underextraction conditions of 10
min, 30 °C and a ratio of 1:30 (w/w). The effective yield of
β-carotene extracted using 100 %hexane was higher than using the
microemulsion, i.e. 41.16±0.12 μg/g for UCP and 47.32±0.10 μg/g for
PCP. The β-carotene content of PCP and UCP were also extracted
with100 % Capmul oil at a fixed extraction time, temperature
andratio of 60min, 60 °C and 1:30 (w/w), respectively. The
valuesobtained from the oil extraction were 25.10±0.11 μg/g forUCP
and 29.38±0.13 μg/g for PCP. The extraction efficien-cies using
hexane and pure oil were lower than for themicroemulsion containing
only 30 % oil, despite containing>99 % hydrophobic compounds
which is more favourable todissolve carotenoids. The reasons for
this may be due to (i)rehydration of carrot cells taking place
during extraction withmicroemulsions which expanded the cells and
increased theintracellular pressure, consequently improving the
extractionefficiency; (ii) the microemulsion having a very small
particlesize (0.05) when the temperature was increased to70 °C. It
was previously reported that temperatures higherthan 65 °C might
cause an increase in β-carotene degradation(Baysal et al. 2000).
Thus, the temperature range of 50–70 °Cwas selected for further
study and setting of the boundaries forthe Box–Behnken experimental
design.
The effects of extraction time (10, 20, 30, 60, 90 and110 min)
on β-carotene content in the microemulsions were
studied while keeping the extraction temperature and
ratioconstant at 60 °C and 1:30 w/w, respectively. The
resultsshowed that the concentration of β-carotene was
significantlyhigher (P
-
27 (33) for a single replicate of a full factorial design
whichreduces to 15 using a Box–Behnken experimental design.
Theresults of this limited number of experiments provided
astatistical model that was used to identify trends in
highconcentration of β-carotene and low PDI and particle sizefor
the extraction process. The data obtained in the Box–Behnken
experiment was used to fit a second-order polyno-mial surface
equation using three independent variables (Xvalues) for each of
the three response variables (Yvalues) asdescribed by Eqs. (4) to
(6):
Y 1 ¼ 21:413þ 2:267X 1 þ 2:412X 2−3:685X 3−2:545X 12−4:765X
22−2:225X 32−0:727X 1X 2−2:927X 1X 3−4:012X 2X 3
ð4Þ
Y 2 ¼ 0:320þ 0:006X 1−0:0125X 2−0:196X 3 þ 0:017X 12 þ 0:035X
22þ0:187X 32 þ 0:072X 1X 2 þ 0:020X 1X 3−0:0175X 2X 3
ð5Þ
Y 3 ¼ 152:9þ 96:04X 1 þ 10:45X 2−278:42X 3 þ 59:38X 12 þ 65:88X
22þ206:13X 32 þ 61:58X 1X 2−81:17X 1X 3 þ 64:31X 2X 3
ð6Þ
Table 2 shows that the models can adequately predict
themicroemulsion properties (i.e. the predicted values of
β-carotene content, PDI and particle size) compared to theresulting
characteristics of the experimentally preparedmicroemulsion (i.e.
the experimental values), for thepredefined extraction time,
extraction temperature and PCP/microemulsion ratio
combinations.
Sen and Swaminathan (2004) reported that analysis ofvariance is
essential to test the significance of the model. Byapplying ANOVA
to the three regression Eqs. (4) to (6), themodels were found to be
significant (P0.1) if the model fits the dataadequately (Cho et al.
2013). The P values of the lack-of-fitfor the models were 0.62,
0.31 and 0.36 for β-carotene, PDIand particle size, respectively.
Insignificant lack of fit, togetherwith the high values of R2,
Ra
2 and R indicated that thequadratic equation was capable of
representing the systemunder the given experimental domain.
The significance of the model and each coefficient wasdetermined
using F test values and corresponding P valuesthat are listed in
Table 3. Larger F values and smaller P valuesindicate that the
parameters in the regression model are moreimportant (Yetilmezsoy
et al. 2009; Khajeh 2011). In thisstudy, taking the F values
(58.07, 94.78 and 4,308.26 for β-carotene, PDI and particle size,
respectively) into consider-ation, all models were significant
(Table 3). For β-carotenecontent, the results showed that the
linear-term and quadratic-term coefficient effects of time,
temperature and ratio weresignificant, as was evident from their
respective F value and Pvalues (Table 3). Based on the mean squares
obtained from theANOVA, the linear-term coefficient of extraction
ratio (X3),quadratic-term coefficient of temperature (X2
2) and the inter-action effect of temperature and ratio (X2X3)
of β-caroteneshowed the very high level of significance compared to
other
3342 Food Bioprocess Technol (2014) 7:3336–3348
-
terms. However, no significant difference (P>0.05) was
ob-served on the interaction between time and
temperature(X1X2).
For PDI, the linear-term (X3), quadratic-term (X22, X3
2)coefficients and interaction of time and temperature
(X1X2)were significant. On other hand, all other terms were
notsignificant. Thus, the extraction ratio could be considered
asthe most significant single parameter. The particle size
modelshows that all linear-term coefficients (X1, X2, X3),
quadratic-term coefficients (X1
2, X22, X3
2) and interaction parameters(X1X2, X1X3, X2X3) were significant
(P
-
pomace absorbing the o/w microemulsion containing 50 %water by
simple diffusion and osmosis through the porescreated during PEF
treatment. This would lead to expansionof the cells and
consequently β-carotene could be solubilizedin the oil matrix and
released by increasing the extraction time.The maximum β-carotene
content of 27.2 μg/g was obtained
at 107.0 min, 62.0 °C and 1:30 (w/w) for X1, X2 and
X3,respectively.
PDI indicates the size distribution of the microemulsiondiscrete
phase, which was significantly affected by the time,temperature and
ratio of the extraction. The response surfacegraph (Fig. 2d–f)
showed that PDI decreased by increasing the
Table 3 Regression coefficientsof the predicted quadratic
poly-nomial models
df degrees of freedom, R2 deter-mination coefficient, Ra
2 adjusteddetermination coefficient, PDIpolydispersity index
Source Sum of squares df Mean square F value P value R2 Ra2
For β-carotene
Model 409.44 9 45.49 58.07 0.00 99.05 % 97.35 %
X1 41.13 1 41.13 52.50 0.00
X2 46.56 1 46.56 59.43 0.00
X3 108.63 1 108.63 138.66 0.00
X12 15.63 1 23.92 30.53 0.00
X22 78.40 1 83.85 107.02 0.00
X32 18.29 1 18.29 23.34 0.01
X1X2 2.12 1 2.12 2.70 0.16
X1X3 34.28 1 34.28 43.76 0.00
X2X3 64.40 1 64.40 82.20 0.00
Residual 3.92 5 0.78
Lack-of-fit 2.07 3 0.69 0.75 0.62
Pure error 1.84 2 0.92
For PDI
Model 0.465 9 0.052 94.78 0.000 99.42 % 98.37 %
X1 0.000 1 0.000 0.570 0.483
X2 0.001 1 0.001 2.290 0.190
X3 0.308 1 0.308 565.34 0.000
X12 0.000 1 0.001 2.07 0.209
X22 0.002 1 0.005 8.30 0.035
X32 0.130 1 0.130 238.18 0.000
X1X2 0.021 1 0.021 38.58 0.002
X1X3 0.002 1 0.002 2.94 0.147
X2X3 0.001 1 0.001 2.25 0.194
Residual 0.003 5 0.001
Lack-of-fit 0.002 3 0.001 2.360 0.31
Pure error 0.001 2 0.000
For particle size
Model 925,007 9 102,779 4,308.26 0.000 99.99 % 99.96 %
X1 73,793 1 73,793 3,093.26 0.000
X2 873 1 873 36.59 0.002
X3 620,147 1 620,147 25,995.27 0.000
X12 5,960 1 13,021 545.8 0.000
X22 9,293 1 16,023 671.67 0.000
X32 156,879 1 156,879 6,576.03 0.000
X1X2 15,168 1 15,168 635.83 0.000
X1X3 26,351 1 26,351 1,104.58 0.000
X2X3 16,542 1 16,542 693.4 0.000
Residual 119 5 24
Lack-of-fit 89 3 30 1.93 0.36
Pure error 31 2 15
3344 Food Bioprocess Technol (2014) 7:3336–3348
-
extraction temperature and ratio whereas it increased by
in-creasing the extraction time. The minimum predicted PDI of0.20
was obtained at 10 min, 70 °C and 1:79 (w/w) for X1, X2and X3,
respectively. In addition to PDI, particle size is also acrucial
characteristic of microemulsion because it influencesthe bioactive
release rate and absorption (Cho et al. 2013).Similar to PDI, the
response surfaces (Fig. 2g–i) show thatparticle size increases with
increasing the extraction tempera-ture and time, whereas it
decreases when increasing the ex-traction ratio. The results
clearly show that appropriate lowerextraction temperatures are
important in controlling particlesize. The increase in particle
size after increasing the extrac-tion temperature and time might be
due to the agglomerationof particles during extraction. The minimum
predicted particlesize of 46.8 nm could be obtained at 47.4 min,
44.5 °C and1:80 (w/w) for X1, X2 and X3, respectively.
All of the three responses should be evaluated in
theoptimization of β-carotene microemulsion. However, it isalmost
impossible to optimize all the objectives at thesame time because
they do not coincide with each otherand conflict may occur between
them. The optimumcondition obtained in one response may have
contraryeffect on another response (Cho et al. 2013). Therefore,in
order to find the best compromising formulation forall responses,
the optimized formulation was achievedusing a desirability function
to satisfy the desired goalssimultaneously: maximized β-carotene
content and min-imized PDI and particle size. Preliminary
investigationshowed that the optimum points for Y1, Y2 and Y3(Fig.
2) did not fall exactly in the same region. Thus,the multi-response
optimization process was transformedinto a single response using
mathematical calculation
A B C
D F
G HI
E
Fig. 2 Response surface graph for β-carotene (a–c), PDI (d–f)
and particle size (g–i) from pulsed electric field-treated carrot
pomace in microemulsionsas a function of temperature, time and
ratio of carrot pomace to microemulsion. Variables not shown in
each plot were kept constant at the centre levels
Table 4 Droplet size, polydis-persity index and zeta potential
ofoptimized condition after storagefor 4 weeks at room
temperature(25 °C)
Predicted value Experimental results
0 weeks 2 weeks 4 weeks
β-carotene content (μg/g) 19.62 19.81±0.13 19.79±0.11
19.76±0.11
Particle size (nm) 74.02 74.90±3.60 73.80±4.0 73.80±3.10
Polydispersity index 0.27 0.27±0.01 0.27±0.01 0.27±0.02
Zeta potential (mV) – −38±7 −37±6 −38±5
Food Bioprocess Technol (2014) 7:3336–3348 3345
-
(Cho et al. 2013). The results show that the optimizedlevels of
X1, X2 and X3 were 49.4 min, 52.2 °C and 1:70w/w, respectively, and
the predicted values of Y1, Y2 andY3 were 19.6 μg/g, 0.27 and 74.0
nm, respectively. Theindividual desirability of β-carotene, PDI and
particlesize was 0.91247, 1 and 1, and the combined desirabilitywas
0.96993. To verify the precision of the model, trip-licate
confirmatory samples were prepared under opti-mum conditions and
the results were compared with theoptimized model.
Characterization of Microemulsion Formulation AfterExtraction
Under Optimal Conditions
To validate the adequacy of the model, additional
experimentswere carried out to verify the predicted values. The
stability ofmicroemulsion after extraction was also monitored after
2 and4 weeks storage as shown in Table 4. Further
examinationsincluding measuring the particle size, PDI and zeta
potential,were carried out to anticipate the stability of the
β-carotene-loaded o/w microemulsion (Table 4). The results indicate
thatthe predicted values were very similar to the
experimentalresults, demonstrating that coupling with a
desirability func-tion could be an efficient approach to optimize
the extractionof β-carotene. In the period of the stability study
(2 and4 weeks), no phase separation and no considerable change
inβ-carotene content, PDI, particle size or zeta potential
wereobserved, indicating that the microemulsion was stable
duringstorage. The zeta potential under the optimized conditions
was−37.8±6.5 mV which confirmed the previous reports that±30 mVof
potential would be sufficient to prevent dropletcoalescence (Cho et
al. 2013).
The encapsulation efficiency of β-carotene-loadedmicroemulsions
prepared under optimized condition wasevaluated by determining the
difference between the totalamount of β-carotene before and after
filtration. The recoveryobtained was 98.2±0.5 %, indicating high
loading efficiency.This demonstrates that most of the β-carotene
was entrappedwithin droplets of the microemulsion. This result was
inagreement with other studies reporting the advantage of usinga
microemulsion technique for encapsulation of drug andbioactive
compounds to achieve a high encapsulation efficien-cy (Qu et al.
2014; Yi et al. 2012).
The appearance of the blank and β-carotene-loaded o/wwas
assessed by TEM using a negative-staining technique.TEM images are
presented in Fig. 3 and shows that thesubmicron-sized droplets were
well dispersed without anyaggregation or cluster formation,
spherical in shape and ap-preciably homogeneous in size. The sizes
are in good agree-ment with the light scattering measurements; a
significantincrease in droplet size was observed in the absence
(64.7±1.1, PDI of 0.30±0.02) and presence (74.9±4.6, PDI of
0.27±0.01) of β-carotene.
Conclusions
This is the first reported work on the use of
oil-in-watermicroemulsions as media for the effective extraction of
β-carotene from carrot pomace. PEF treatment improved
theextractability of β-carotene compared to untreated carrotpomace.
The effects of three independent variables (extractiontime,
temperature and carrot pomace to microemulsion ratio)on β-carotene
content, PDI and particle size were evaluatedby a Box–Behnken
design, and optimized using a desirabilityfunction. The models
showed that the extraction ratio was themost significant single
parameter which influenced the extrac-tion of β-carotene, PDI and
particle size, followed by extrac-tion temperature and extraction
time. Under optimum condi-tions, the desirable observed responses
were close to thepredicted values. The suitable extraction
conditions were ob-tained at extraction time 49.4 min, extraction
temperature
Fig. 3 Transmission electron micrographs of blank microemulsion
drop-lets (a) and β-carotene-loaded microemulsion prepared under
optimizedconditions (b). Scale bar 1,000 nm and magnification
×13,500
3346 Food Bioprocess Technol (2014) 7:3336–3348
-
52.2 °C and extraction ratio 1:70 (w/w). Under these
condi-tions, the response variables were predicted to be 19.6
μg/g,0.27 and 74 nm for β-carotene content, PDI and particle
size,respectively. This transparent β-carotene-loadedmicroemulsion
could be used as a promising vehicle forfortification of
transparent food, beverage and pharmaceuticalproducts.
Acknowledgments This research was part of the New
Zealand-Koreajoint grant programme funded by the Ministry of
Business, Innovation,and Employment (MBIE). The authors gratefully
acknowledge RichardEasingwood for his help with the electron
microscopy techniques andUniversity of Otago for PhD Scholarship
for Shahin Roohinejad.
References
Arvanitoyannis, I. S. (2009). Encapsulation and controlled
release tech-nologies in food systems. [Article]. International
Journal of FoodScience & Technology, 44(7), 1462.
Bayraktar, E. (2001). Response surface optimization of the
separation ofdl-tryptophan using an emulsion liquid membrane.
ProcessBiochemistry, 37(2), 169–175.
Baysal, T., Ersus, S., & Starmans, D. A. J. (2000).
Supercritical CO2extraction of β-carotene and lycopene from tomato
paste waste.Journal of Agricultural and Food Chemistry, 48(11),
5507–5511.
Budker, V. G., Slattum, P. M., Monahan, S. D., & Wolff, J.
A. (2002).Entrapment and condensation of DNA in neutral reverse
micelles.Biophysical Journal, 82(3), 1570–1579.
Cai, Y.-Z., Sun, M., & Corke, H. (2005). Characterization
and applicationof betalain pigments from plants of the
Amaranthaceae. Trends inFood Science & Technology, 16(9),
370–376.
Chen, G., & Djuric, Z. (2001). Carotenoids are degraded by
free radicalsbut do not affect lipid peroxidation in unilamellar
liposomes underdifferent oxygen tensions. FEBS Letters, 505(1),
151–154.
Cho, H. J., Lee, D. W., Marasini, N., Poudel, B. K., Kim, J.
H.,Ramasamy, T. , e t a l . (2013) . Opt imiza t ion of se l f
-microemulsifying drug delivery system for telmisartan using
Box–Behnken design and desirability function. Journal of Pharmacy
andPharmacology, 65(10), 1440–1450.
Cortez, E. V., Pessoa, A., Jr., Felipe, M., Roberto, I. C.,
& Vitolo, M.(2004). Liquid–liquid extraction of xylitol
dehydrogenase fromCandida guilliermondii homogenate by reversed
micelles. Journalof Chromatography B, 807(1), 55–60.
Dantas, T. N. C., Neto, A. A. D., Moura,M. C. P. A., Neto, E. L.
B., Forte,K. R., & Leite, R. H. L. (2003). Heavy metals
extraction bymicroemulsions. Water Research, 37(11), 2709–2717.
Donsì, F., Ferrari, G., & Pataro, G. (2010). Applications of
pulsed electricfield treatments for the enhancement of mass
transfer from vegetabletissue. Food Engineering Reviews, 2(2),
109–130.
Ehen, H. M., & Meyers, S. P. (1982). Extraction of
astaxanthin pigmentfrom crawfish waste using a soy oil process.
Journal of FoodScience, 47(3), 892–896.
El Darra, N., Grimi, N., Maroun, R., Louka, N., & Vorobiev,
E. (2013).Pulsed electric field, ultrasound, and thermal
pretreatments for betterphenolic extraction during red
fermentation. European FoodResearch and Technology, 236(1),
47–56.
Ersus, S., & Barrett, D. M. (2010). Determination of
membrane integrityin onion tissues treated by pulsed electric
fields: use of microscopicimages and ion leakage measurements.
Innovative Food Science &Emerging Technologies, 11(4),
598–603.
Fan, Y., Ma, L., Zhang, W., Wang, J., Chen, Y., Gao, Y., et al.
(2014). Thedesign of propolis flavone microemulsion and its effect
on enhanc-ing the immunity and antioxidant activity in mice.
InternationalJournal of Biological Macromolecules, 65, 200–207.
Flanagan, J., & Singh, H. (2006). Microemulsions: a
potential deliverysystem for bioactives in food. Critical Reviews
in Food Science andNutrition, 46(3), 221–237.
Fratianni, A., Cinquanta, L., & Panfili, G. (2010).
Degradation of carot-enoids in orange juice during microwave
heating. LWT - FoodScience and Technology, 43(6), 867–871.
Gachovska, T., Cassada, D., Subbiah, J., Hanna, M., Thippareddi,
H., &Snow, D. (2010). Enhanced anthocyanin extraction from red
cab-bage using pulsed electric field processing. Journal of Food
Science,75(6), E323–E329.
Grimi, N., Praporscic, I., Lebovka, N., & Vorobiev, E.
(2007). Selectiveextraction from carrot slices by pressing and
washing enhanced bypulsed electric fields. Separation and
Purification Technology,58(2), 267–273.
Hedrén, E., Diaz, V., & Svanberg, U. (2002). Estimation of
carotenoidaccessibility from carrots determined by an in vitro
digestion meth-od. European Journal of Clinical Nutrition, 56(5),
425–430.
Hiranvarachat, B., & Devahastin, S. (2014). Enhancement
ofmicrowave-assisted extraction via intermittent radiation:
ex-traction of carotenoids from carrot peels. Journal of
FoodEngineering, 126, 17–26.
Illés, V., Daood, H. G., Biacs, P. A., Gnayfeed, M. H., &
Mészáros, B.(1999). Supercritical CO2 and subcritical propane
extraction of spicered pepper oil with special regard to carotenoid
and tocopherolcontent. Journal of Chromatographic Science, 37(9),
345–352.
Khajeh, M. (2011). Response surface modelling of lead
pre-concentrationfrom food samples by miniaturised homogenous
liquid–liquid sol-vent extraction: Box–Behnken design. Food
Chemistry, 129(4),1832–1838.
Kritchevsky, S. B. (1999). β-Carotene, carotenoids and the
pre-vention of coronary heart disease. The Journal of
Nutrition,129(1), 5–8.
Lemmens, L., Van Buggenhout, S., Oey, I., Van Loey, A., &
Hendrickx,M. (2009). Towards a better understanding of the
relationship be-tween the β-carotene in vitro bio-accessibility and
pectin structuralchanges: a case study on carrots. Food Research
International,42(9), 1323–1330.
Liu, H.-L., Lan, Y.-W., & Cheng, Y.-C. (2004). Optimal
production ofsulphuric acid by Thiobacillus thiooxidans using
response surfacemethodology. Process Biochemistry, 39(12),
1953–1961.
López, N., Puértolas, E., Condón, S., Raso, J., & Alvarez,
I. (2009).Enhancement of the extraction of betanine from red
beetroot bypulsed electric fields. Journal of Food Engineering,
90(1), 60–66.
Mustafa, A., Trevino, L. M., & Turner, C. (2012).
Pressurized hot ethanolextraction of carotenoids from carrot
by-products.Molecules (Basel,Switzerland), 17(2), 1809–1818.
Ofori-Boateng, C., & Lee, K. T. (2013). Response surface
optimization ofultrasonic-assisted extraction of carotenoids from
oil palm (Elaeisguineensis Jacq.) fronds. Food Science and
Nutrition, 1(3), 209–221.
Patel, M. R., Patel, R. B., Parikh, J. R., Solanki, A. B., &
Patel, B. G.(2009). Effect of formulation components on the in
vitro permeationof microemulsion drug delivery system of
fluconazole. AAPSPharmSciTech, 10(3), 917–923.
Paul, B. K., & Moulik, S. P. (2001). Uses and applications
ofmicroemulsions. Current Science, 80(8), 990–1001.
Puértolas, E., Cregenzán, O., Luengo, E., Álvarez, I., &
Raso, J.(2013). Pulsed-electric-field-assisted extraction of
anthocya-nins from purple-fleshed potato. Food Chemistry,
136(3–4),1330–1336.
Qu, D., He, J., Liu, C., Zhou, J., & Chen, Y. (2014).
Triterpene-loadedmicroemulsion using Coix lacryma-jobi seed extract
as oil phase for
Food Bioprocess Technol (2014) 7:3336–3348 3347
-
enhanced antitumor efficacy: preparation and in vivo
evaluation.International Journal of Nanomedicine, 9, 109.
Reiter, M., Neidhart, S., & Carle, R. (2003). Sedimentation
behaviour andturbidity of carrot juices in relation to the
characteristics of theircloud particles. Journal of the Science of
Food and Agriculture,83(8), 745–751.
Roohinejad, S., Oey, I., & Everett, D. W. (2014). Effect of
pulsed electricfield processing on carotenoid extractability of
carrot purée.International Journal of Food Sciences and Nutrition.
doi:10.1111/ijfs.12510.
Sachindra, N. M., &Mahendrakar, N. S. (2005). Process
optimization forextraction of carotenoids from shrimp waste with
vegetable oils.Bioresource Technology, 96(10), 1195–1200.
Sanchez-Moreno, C., Plaza, L., Elez-Martinez, P., De Ancos, B.,
Martin-Belloso, O., & Cano, M. P. (2005). Impact of high
pressure andpulsed electric fields on bioactive compounds and
antioxidant ac-tivity of orange juice in comparison with
traditional thermal pro-cessing. Journal of Agricultural and Food
Chemistry, 53(11), 4403–4409.
Schmidts, T., Dobler, D., von den Hoff, S., Schlupp, P., Garn,
H., &Runkel, F. (2011). Protective effect of drug delivery
systems againstthe enzymatic degradation of dermally applied
DNAzyme.International Journal of Pharmaceutics, 410(1–2),
75–82.
Sen, R., & Swaminathan, T. (2004). Response surface modeling
andoptimization to elucidate and analyze the effects of inoculum
ageand size on surfactin production. Biochemical Engineering
Journal,21(2), 141–148.
Spernath, A., Yaghmur, A., Aserin, A., Hoffman, R. E., &
Garti, N.(2002). Food-grade microemulsions based on nonionic
emulsifiers:media to enhance lycopene solubilization. Journal of
Agriculturaland Food Chemistry, 50(23), 6917–6922.
Sun, Y., Liu, D., Chen, J., Ye, X., & Yu, D. (2011). Effects
of differentfactors of ultrasound treatment on the extraction yield
of the all-
trans-beta-carotene from citrus peels. Ultrasonics
Sonochemistry,18(1), 243–249.
Talcott, S. T., & Howard, L. R. (1999). Phenolic
autoxidation is respon-sible for color degradation in processed
carrot puree. Journal ofAgricultural and Food Chemistry, 47(5),
2109–2115.
Talegaonkar, S., Azeem, A., Ahmad, F. J., Khar, R. K., Pathan,
S. A., &Khan, Z. I. (2008). Microemulsions: a novel approach to
enhanceddrug delivery. Recent Patents on Drug Delivery &
Formulation,2(3), 238–257.
Tsubaki, S., Sakamoto, M., & Azuma, J.-i. (2010).
Microwave-assisted extraction of phenolic compounds from tea
residuesunder autohydrolytic conditions. Food Chemistry,
123(4),1255–1258.
Van Poppel, G., & Goldbohm, R. A. (1995). Epidemiologic
evidence forbeta-carotene and cancer prevention. The American
Journal ofClinical Nutrition, 62(6), 1393S–1402S.
Wright, A. J., Pietrangelo, C., & MacNaughton, A. (2008).
Influence ofsimulated upper intestinal parameters on the efficiency
of betacarotene micellarisation using an in vitro model of
digestion. FoodChemistry, 107(3), 1253–1260.
Yetilmezsoy, K., Demirel, S., & Vanderbei, R. J. (2009).
Responsesurface modeling of Pb(II) removal from aqueous solution
byPistacia vera L.: Box–Behnken experimental design. Journal
ofHazardous Materials, 171(1–3), 551–562.
Yi, C., Zhong, H., Tong, S., Cao, X., Firempong, C. K., Liu, H.,
et al.(2012). Enhanced oral bioavailability of a
sterol-loadedmicroemulsion formulation of Flammulina velutipes, a
potentialantitumor drug. International Journal of Nanomedicine, 7,
5067–5078.
Zhao, L., Zhao, G., Chen, F., Wang, Z., Wu, J., & Hu, X.
(2006). Differenteffects of microwave and ultrasound on the
stability of (all-E)-astaxanthin. Journal of Agricultural and Food
Chemistry, 54(21),8346–8351.
3348 Food Bioprocess Technol (2014) 7:3336–3348
http://dx.doi.org/10.1111/ijfs.12510http://dx.doi.org/10.1111/ijfs.12510
Evaluating...AbstractIntroductionMaterials and
MethodsReagentsPreparation of Pulsed Electric Field-Treated Carrot
PomaceExperimental Designβ-Carotene Extraction Using a
MicroemulsionMicroemulsion Droplet Size, PDI and Zeta Potential
AnalysisMicroemulsion StabilityEncapsulation
EfficiencyDetermination of Carotenoid Content in the
MicroemulsionTransmission Electron MicroscopyStatistical
Analysis
Results and DiscussionEvaluation of MethodEffect of Pulsed
Electric Field on Release of β-CaroteneSelection of Extraction
Conditions
Optimization of MethodStatistical Evaluation of the Developed
Model for β-Carotene ExtractionEffects of Extraction Parameters on
ResponsesCharacterization of Microemulsion Formulation After
Extraction Under Optimal Conditions
ConclusionsReferences