Food Hydrocolloids 38 (2014) 172e179

Contents lists available at ScienceDirect

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

Comparative study of encapsulation of vitamins with native

and modified soy protein
A. Nesterenko a, b, I. Alric a, b, F. Silvestre a, b, V. Durrieu a, b, *
a
Université de Toulouse, INP-ENSIACET, LCA (Laboratoire de Chimie Agro-Industrielle), F-31030 Toulouse, France
b
INRA, UMR 1010 CAI, F-31030 Toulouse, France

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

Article history: Microencapsulation of hydrophobic (a-tocopherol) and hydrophilic (ascorbic acid) vitamins by native
Received 15 July 2013 (non-modified) and modified soy protein isolate (SPI) was carried out using a spray-drying technique.
Accepted 2 December 2013 Proteins’ functional properties were modified by acylation and cationization reactions in aqueous
alkaline media. The results obtained demonstrated that SPI modification resulted in decreased emulsion
Keywords: droplet size and viscosity. All preparations with ascorbic acid (AA) had lower viscosity and microparticle
Microencapsulation
size than those with a-tocopherol (T). Moreover, grafting of fatty acid chains to SPI by acylation improved
Soy protein
its amphiphilic character and affinity with hydrophobic substances. Thus, the microencapsulation effi-
a-Tocopherol
Ascorbic acid
ciency of T was increased from 79.7% to 94.8% and the microencapsulation efficiency of AA was reduced
Acylation from 91.8% to 57.3% compared to native SPI. Conversely, attachment of quaternary ammonium cationic
Cationization groups to proteinic chains by cationization, increased SPI solubility and favored the AA microencapsu-
lation. This study illustrated that an appropriate modification of SPI can improve the microencapsulation
efficiency of suitable active cores.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction active core substance from the surrounding environment within a

wall or matrix material. This technique offers benefits for protec-
Over the past decades, environmental requirements have tion of sensitive compounds, controlled release of the core agent,
become of great importance. In order to replace synthetic polymers masking of unpleasant taste and odor of the substances or trans-
and animal derived products, there is an increasing interest in the formation of liquid core into solid powder. Different processes
industrial use of renewable resources and development of naturally could be used to produce microparticles: spray-drying, spray-
occurring materials for new applications. Natural polymers such as cooling/chilling, supercritical fluid expansion, fluidized bed, gela-
vegetable proteins have attracted considerable research activities tion, solvent evaporation, coacervation and extrusion (Augustin &
because of their availability, biodegradability, renewable character Hemar, 2009; Dubey, Shami, & Bhasker, 2009; Gouin, 2004).
and various interesting functional properties. Among them, pro- Spray-drying consists of the conversion of a liquid preparation
teins extracted from vegetable seeds (soybean, pea, barley, wheat, (containing wall and core material) into a solid powder of micro-
rice, oat, sunflower) have been reported as having good emulsifying particles using a stream of heated air. This technology, widely used
and foaming capacities, water solubility, amphiphilic and film- in industry, is commonly employed for microencapsulation of
forming properties (Nunes, Batista, Raymundo, Alves, & Sousa, various active substances with a vegetable protein matrix. Among
2003). vegetable proteins, soy proteins, pea proteins, wheat proteins and
Due to their good physico-chemical properties, vegetable pro- barley proteins had already demonstrated their effectiveness as
teins represent a highly suitable wall-forming material for micro- carrier materials in microencapsulation by spray-drying
encapsulation of active components to use in the food industry, (Nesterenko, Alric, Silvestre, et al., 2013).
pharmaceutics and cosmetics (Nesterenko, Alric, Silvestre, & Soy proteins represent an important component of soy bean
Durrieu, 2013). Microencapsulation allows the isolation of the seeds (35e40%). Two fractions are mainly present in extracted soy
proteins: glycinin (11S globulin) and conglycinin (7S globulin). Soy
protein isolate (SPI) showed interesting physico-chemical proper-
ties in particular gelling, emulsifying, fat-absorbing and water
* Corresponding author. INRA, UMR 1010 CAI, F-31030 Toulouse, France. binding (Caillard, Remondetto, & Subirade, 2009; Hua, Cui, Wang,
E-mail address: vanessa.durrieu@ensiacet.fr (V. Durrieu). Mine, & Poysa, 2005; Nunes et al., 2003). The use of SPI as wall

0268-005X/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodhyd.2013.12.011
 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179 173

material in microencapsulation by spray-drying had been reported dodecanoyl chloride, glycidyltrimethylammonium chloride, cyclo-
by various authors. This natural polymer showed a high efficiency hexane (HPLC grade), iodine and sodium thiosulfate were pur-
for coating different active substances: orange oil (Kim, Morr, & chased from Sigma (Saint-Quentin Fallavier, France).
Schenz, 1996), fish oil (Augustin, Sanguansri, & Bode, 2006), stea-
rin/palm oil (Rusli, Sanguansri, & Augustin, 2006), phospholipid 2.2. SPI modifications
(Yu, Wang, Yao, & Liu, 2007), flavors (Charve & Reineccius, 2009),
casein hydrolysate (Favaro-Trindade, Santana, Monterrey-Quintero, The acylation reaction was carried out on SPI using dodecanoyl
Trindade, & Netto, 2010; Ortiz, Mauri, Monterrey-Quintero, & chloride (DDC) following the SchotteneBaumann reaction as
Trindade, 2009), paprika oleoresin (Rascon, Beristain, Garcie, & described previously (Nesterenko, Alric, Silvestre, & Durrieu, 2012).
Salgado, 2011) and soy oil (Tang & Li, 2013). The molar ratio DDC/NH2 of protein used for the reaction was 0.5/1
It is widely accepted that the antioxidant properties of a- and the sample obtained was named SPI-A.
tocopherol (vitamin E) and ascorbic acid (vitamin C) are responsible The SPI cationization reaction was carried out in aqueous solu-
in part for their biological activity (Packer, Slater, & Willson, 1979). tion (5% w/w) at 40
C or 70
C using glycidyltrimethylammonium
Nevertheless, environmental factors, such as oxygen, temperature, chloride (GTMAC). When the SPI solution reached reaction tem-
moisture and UV affect the stability of these compounds and perature, pH was adjusted to 10.0 with 4 M NaOH and GTMAC was
involve their deterioration. Microencapsulation could be an effi- added (molar ratios GTMAC/NH2 were 1, 2 or 4). The pH of the
cient method for the protection and stabilization of a-tocopherol solution was maintained at 10.0 during the 1 h reaction period. The
(T) and ascorbic acid (AA). However, the nature of the wall matrix reaction was ended by adjusting the pH of the solution to 7.0 using
particularly affects the degree of protection of the active core, the 4 M HCl. The mixture of cationized SPI was cooled, freeze-dried at
microparticles’ stability and the retention efficiency. 20 Pa (Cryo-Rivoire equipment, Cryonext, Saint Gely du Fesc,
Modifiable character is one of the important advantages of France) and stored at 4
C. Samples obtained were named SPI-C.
proteins. Modification of proteinic chains leads to changes in the The degree of cationization (DC) was evaluated using the o-phtal-
properties and behavior of this natural polymer and diversification dialdehyde method (OPA) (Church, Swaisgood, Porter, & Catignani,
of protein functionalities. In microencapsulation, functionalization 1983; Goodno, Swaisgood, & Catignani, 1981) and defined as
of proteinic chains makes it possible to obtain microparticles with follows:
new properties, different from those obtained with other wall
materials. ðn0  nm Þ
DCð%Þ ¼  100 (1)
Grafting of fatty acid chains to proteins by acylation is well n0
known in order to enhance their hydrophobicity, surface-active
functionality and emulsifying capacity (Matemu, Kayahara, where n0 is the molar quantity of amino groups per gram of native
Murasawa, Katayama, & Nakamura, 2011; Wong, Nakamura, & SPI, and nm the molar quantity of amino groups per gram of cat-
Kitts, 2006). In fact, the incorporation of hydrocarbon chains (hy- ionized SPI.
drophobic part) into the protein macromolecules (hydrophilic
part) allows the creation of amphiphilic structures with improved 2.3. SPI solubility profiles
surface activity (Rondel, Alric, Mouloungui, Blanco, & Silvestre,
2009). On the other hand, the introduction of quaternary ammo- Native SPI solubility was compared to the cationized sample
nium groups to polysaccharides (Channasanon, Graisuwan, (SPI-C) and the blank sample (SPI-Cblank treated under cationization
Kiatkamjornwong, & Hoven, 2007; Wang et al., 2012) or to ani- conditions without GTMAC). Solubility profiles of SPI were deter-
mal derived proteins (Kiick-Fischer & Tirrell, 1998; Zohuriaan- mined as described in a previous study (Nesterenko et al., 2012).
Mehr, Pourjavadi, Salimi, & Kurdtabar, 2009) by cationization is Briefly, protein mixtures in deionized water (5% w/w) were pre-
used to enhance their solubility, antibacterial properties as well as pared at different pH values and stirred at 70
C for 1 h. Suspensions
their hydrophilic properties (water absorption and swelling ca- were centrifuged at 10,000  g for 15 min (Sigma Laborzentrifugen,
pacity). Nevertheless, there is no data in the literature dealing with Osterode, Germany). The soluble protein fraction in the superna-
cationization of vegetable proteins. Both acylation and cationiza- tant was analyzed using the Kjeldahl method and solubility (S%, w/
tion reactions could be suggested as an effective way to obtain SPI w) was calculated from the following equation:
with defined characteristics.
Therefore, the objective of this work was to study the influ- protein content in the supernatant
ence of SPI modification by acylation and cationization, on the Sð%Þ ¼  100 (2)
total protein content in solution
microencapsulation of hydrophobic (T) and hydrophilic (AA) vi-
tamins by the spray-drying technique. In the context of “green”
chemistry (Ga1uszkaa, Migaszewskia, & Namiesnik, 2013), modi-
fication reactions were carried out without any use of organic 2.4. Microencapsulation by spray-drying
solvents and chemical catalysts. The effect of SPI modifications on
both solution/emulsion and microparticle properties was also Protein based microparticles were prepared using a two-step
investigated. procedure. An aqueous solution of protein (native or modified)
was mixed with active core material. Then a liquid preparation
2. Materials and methods (solution or emulsion) was spray-dried to obtain a microparticle
powder.
2.1. Materials
2.4.1. Solution/emulsion preparation
Soy protein isolate (SPI), 90% pure, was purchased from Lustrel The wall material (SPI) was dissolved in deionized water (8% w/
Laboratoires SAS (Saint Jean de Vedas, France). The term ‘native SPI’ w) at 70
C for 1 h under constant mechanical stirring (1000 rpm).
was used in this study for all samples prepared with non-modified In order to allow maximum protein solubilization, the pH of the
commercial soy protein isolate. All other chemicals were of solution was fixed at 10.5. Active material (T or AA) was then mixed
analytical grade. a-Tocopherol, L-ascorbic acid, sodium hydroxide, with SPI solution to obtain the preparation in which the protein/
174 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179

active core ratio was 2/1 (11.5% of total solids). Liquid preparations the solution was measured using a UV Spectrometer (UV-1800,
were homogenized with a high-pressure homogenizer (APV Sys- Shimadzu, Kyoto, Japan) at 298 nm. To determine AA content of the
tems, Albertslund, Denmark) at 50 MPa. The intense mechanical dry microparticles, the AOAC (Association of Official Analytical
force developed during homogenization, contributed to protein Chemists) standard methodology with an iodometric titration
structure modifications, such as unfolding of proteinic chains. Polar procedure was used (AOAC, 2007). Each sample of microparticle
and non-polar regions of the protein were exposed to new envi- was analyzed at least three times.
ronments, which made them more surface-active (Rampon, The microencapsulation process was monitored for both
Riaublanc, Anton, Genot, & McClements, 2003). Preparations with retention efficiency (RE) and load efficiency (LE). RE was defined as
T were named SPI/T, SPI-A/T and SPI-C/T for native, acylated and the percentage of estimated active core content in particles ob-
cationized soy protein respectively. Preparations with AA were tained (Coreexp) over theoretical core content (Coretheo) in initial
named SPI/AA, SPI-A/AA and SPI-C/AA for native, acylated and liquid preparation.
cationized soy protein respectively.
Coreexp
REð%Þ ¼  100 (4)
2.4.2. Solution/emulsion characterization Coretheo
Light scattering and optical microscopy were used to check good The difference between experimental and theoretical values
dispersion of T in the proteinic solution and droplet size uniformity. was caused by active core loss during spray-drying. LE, corre-
Emulsions were analyzed after high-pressure homogenization for sponding to active core content per 100 g of powder, was calculated
1 h. The oil droplet size distribution of homogenized emulsions as:
(SPI/T, SPI-A/T and SPI-C/T) and zeta-potential of protein solutions
were measured using Zetasizer Nano-ZS equipment (Malvern In- mCexp
LEð%Þ ¼  100 (5)
struments, Worcestershire, UK). To avoid multiple scattering ef- mm
fects, liquid preparations were diluted 100 times with deionized
water before measurements. A relative refractive index hoil/ where mCexp is the estimated mass of core in microparticles, and mm
hwater ¼ 1.12 (hoil ¼ 1.49, hwater ¼ 1.33) was used to analyze the data, the mass of the analyzed sample of microparticles (dry matter).
assuming that all droplets were spherical in shape. The volume Particle size distribution was determined by the scattering
particle diameter (D43 or Dv) was taken and used as an indicator of pattern of a transverse laser light using the Scirocco 2000 equip-
the emulsion size. Additionally, emulsions were visualized using an ment (Malvern Instruments, Worcestershire, UK). The volume
Eclipse E600 optical microscope (Nikon, Sendai, Japan), linked to a particle diameter (D43 or Dv) was calculated as the mean of three
digital video camera (DXM1200, Nikon, Sendai, Japan) at a magni- measurements per sample.
fication of 1000. The morphology of the microparticles was observed with a
Apparent viscosity of all liquid preparations after high-pressure scanning electron microscope LEO435VP (LEO Electron microscopy
homogenization was determined at 20
C and shear stress variation Ltd., Cambridge, UK) operated at a voltage of 8 kV. In order to
between 0 and 1 N/m2 for 3 min, using a Rheometer CSL100 (Carri- examine the inner structure of prepared microparticles, the pow-
Med LTD, Dorking, UK) with a 6 cm diameter plate-cone geometry der was first frozen in liquid nitrogen and broken up in a mortar.
and 0.035 rad angle. Liquid preparations were characterized as Samples were deposited on conductive double-sided adhesive tape
Newtonian fluids. and sputter-coated with silver.

2.4.3. Microparticle preparation 2.5. Statistical analysis

Freshly homogenized liquid preparations were spray-dried
using a Mini Spray Dryer B-290 (Büchi, Flawil, Switzerland) un- The experimental data was statistically analyzed using Minitab
der stable process conditions as follows: inlet air temperature at 16 software (State College, USA). A one-way analysis of variance
124  4
C and outlet at 74  4
C, drying air flow rate of 35 m3, (ANOVA) was performed to determine significant differences
spray flow rate of 0.47 m3/h, liquid feed flow rate of (P < 0.05) between the samples. Tukey’s test was adopted as the
0.33  103 m3/h and aspiration of 100%. Microparticles were multiple comparison procedure.
collected from the cyclone collector, shut hermetically in opaque
packaging and stored at 4
C. The spray-drying yield was defined as
follows: 3. Results and discussions

mp 3.1. SPI modifications

Spray  drying yieldð%Þ ¼  100 (3)
mSPIþCore
3.1.1. SPI cationization
where mSPIþCore the initial mass of solids added in liquid prepara- During cationization in alkaline aqueous media, the GTMAC
tion including SPI and active core (T or AA) and mp is the mass of reacted with nucleophilic sites of the protein i.e. primary amino
collected powder (dry matter). The moisture of obtained micro- groups (N-terminal and lysine residues), which were the most
particles was determined with an infrared moisture balance reactive ones. The reagent used (GTMAC) is toxic because of the
(Sartorius, Goettingen, Germany) by drying the sample at 105
C to presence of an epoxy group. However, during cationization at pH
constant weight during 5 min. 10.0 in water, the GTMAC unreacted with protein was hydrolyzed to
(2,3-dihydroxypropyl)trimethylammonium chloride, and this sec-
2.4.4. Microparticle characterization ondary product did not have any toxicity. Indeed, it was used as an
The amount of T retained in microparticles during drying was additive for care products to promote the retention of moisture
determined using UV/VIS spectroscopy (Faria, Mignone, (Baldaro, Pelizzari, Tenconi, & Li Bassi, 2008). The total hydrolysis of
Montenegro, Mercadante, & Borsarelli, 2010). Briefly, about 5 mg GTMAC under the experimental conditions employed for cationi-
of microspheres containing the T to be determined were dissolved zation of SPI, was verified by nuclear magnetic resonance (NMR) of
in 10 mL of cyclohexane. The solution was stirred for 10 min and hydrogen. Thus, cationized SPI was used for microencapsulation
filtered through a 0.2 mm PTFE membrane filter. The absorbance of experiments without previous purification.
 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179 175

Fig. 1. The effect of SPI cationization conditions (temperature and molar ratio GTMAC/NH2) on (A) DC and (B) zeta potential at pH 10.0.

From the results presented on Fig. 1A, the degree of cationiza- shift of the protein isoelectric point into the alkaline range (pH of
tion (DC) of SPI increased with the quantity of GTMAC in the re- 6.0e6.5). After cationization, the number of net positive charges of
action media, and the temperature. Similar phenomena had been the protein was increased and the number of basic NH2 functions
reported in a study dealing with cationization of starch using reduced. This affected the overall balance of acid to basic groups in
GTMAC (Kavaliauskaite, Klimaviciute, & Zemaitaitis, 2008). the protein and resulted in the shift phenomenon.
The highest modification rate (DC ¼ 91.6%) was obtained at a The solubility of SPI-C was higher than the solubility of SPI at pH
molar ratio GTMAC/NH2 of 4/1, after 1 h reaction time at 70
C. To 1e10, which confirmed a significant contribution of polar cationic
confirm the presence of positively charged groups on SPI, zeta- groups on protein affinity with water. At acidic pH, positive charges
potential measurements were made. The zeta potential of native fixed to SPI by cationization increased the repulsion between pro-
SPI at pH 10.0 was 44.4 mV, and as shown on Fig. 1B, the zeta- teinic chains. This favored proteinesolvent interactions and
potential of SPI after cationization had increased. However, values explained the increase in protein solubility. Conversely, in alkaline
obtained were negative because of the high amount of negative net solution (pH > 10), the attraction between cationic ((CH3)3Nþ) and
charges (COO) on the proteinic chains in alkaline media (pH 10.0). carboxylic (COO) groups resulted in the decrease of proteinesol-
After cationization, the majority of primary amino groups (NH2) vent interactions. Thus the lowest protein solubility could be
were replaced by positively charged functions. Thus, the augmen- explained by the formation of more compact structures.
tation in zeta-potential with increase of DC confirmed qualitatively The cationized sample SPI-C (DC of 91.6%) obtained at 70
C and
the presence of positively charged trimethylammonium groups on molar ratio GTMAC/NH2 of 4/1, was used as wall material for
proteinic chains. microencapsulation experiments.
The effect of SPI cationization on its functional properties,
especially the solubility profiles, was studied (Fig. 2). The solubility 3.1.2. SPI acylation
of SPI and SPI-C was compared to a blank sample (SPI-Cblank treated In agreement with our previous report (Nesterenko et al., 2012),
for 1 h at 70
C and pH 10.0 without GTMAC). The blank sample was soy protein (degree of acylation of 33.2%) modified with dodeca-
more soluble than native SPI for all pH values. This observation noyl chloride (DDC) was the more efficient wall material for T
could be related to the partial unfolding of proteinic chains during microencapsulation compared to other modified proteins (hydro-
treatment at alkaline pH and high temperature. On the other hand, lyzed, hydrolyzed and acylated, acylated with octanoyl chloride and
for native SPI and SPI-Cblank samples, the isoelectric point was in a hexadecanoyl chloride). Therefore, SPI acylated with DDC (molar
pH range between 4.5 and 5.0. However, the attachment of posi- ratio DDC/NH2 of 0.5/1) was selected for the present study.
tively charged groups to SPI by cationization resulted in an obvious
3.2. Microencapsulation with native and modified SPI

3.2.1. Effect of SPI modifications on solution/emulsion properties

The preparation of a feed liquid containing wall material and an
active core, was the first step involved in the process of microen-
capsulation by spray-drying. In the case of T encapsulation, the
liquid preparation was an oil-in-water emulsion, whereas it was a
solution in the case of AA coating. The proteinic matrix material
used for T and AA microencapsulation was native (SPI), acylated
(SPI-A) and cationized (SPI-C) soy protein.
The different characterizations of solutions/emulsions and mi-
croparticles are summarized in Table 1. Structural modifications of
proteinic chains affected the viscosity of liquid preparations and the
mean droplet size in emulsions.
Fatty acid hydrophobic chains attached to water-soluble pro-
tein by acylation favored the improvement of the amphiphilic
character of SPI. Moreover, the DDC unreacted with SPI was hy-
drolyzed to sodium dodecanoate having surfactant properties. This
enhancement in surface-active properties of SPI resulted in lower
emulsion viscosity for sample SPI-A/T compared to sample SPI/T. A
Fig. 2. Effect of cationization on the solubility profile of SPI. similar result was reported earlier (Derkatch et al., 2007), and the
176 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179

Table 1
Properties of liquid preparations and spray-dried microparticles based on native and modified SPI.

Measured property Wall/core materialsg

SPI/T SPI/AA SPI-A/T SPI-A/AA SPI-C/T SPI-C/AA

Emulsion droplet size, Dv (mm) 1.1  0.02a

ND 0.7  0.04c
ND 0.9  0.03b
ND
Liquid preparation viscosity (mPa s) 15.0  0.02a 8.9  0.03c 8.0  0.02d 9.8  0.04b 4.1  0.03f 5.8  0.04e
Spray-drying yield (%) 68 81 66 77 61 87
REh (%) 79.7  1.0b 91.8  0.7a 94.8  2.2a 57.3  4.1c 38.3  1.0d 92.3  2.8a
LEi (%) 26.3  0.6b 30.3  0.2a 31.3  0.7a 18.9  1.4c 12.6  0.3d 30.5  0.9a
Mean particle size (mm) 9.3  0.2a 5.2  0.2c 7.7  0.2b 4.8  0.5c 7.6  0.3b 4.9  0.2c
aef
Different letters in the same line indicate a statistical difference between the mean values (P < 0.05).
g
SPI, SPI-A and SPI-C: native, acylated (degree of acylation e 33.2%) and cationized (degree of cationization e 91.6%) soy protein isolate respectively; T: a-tocopherol; AA:
ascorbic acid.
h
RE: retention efficiency of a-tocopherol determined by UV spectroscopy and retention efficiency of ascorbic acid determined by iodometric titration.
i
LE: load efficiency or active core content per 100 g of powder.

phenomenon of viscosity fall was attributed to the changes in the properties of SPI favored the mobility of proteinic chains in aqueous
composition of interface layers in the emulsion. The SPI-A/T media and, presumably, could explain the decreased viscosity for
emulsion (with acylated protein) had a reduced oil droplet size samples SPI-C/T and SPI-C/AA. A decrease in the mean droplet size
compared to SPI/T emulsion (0.7 mm and 1.1 mm respectively). The in the case of T based emulsions was observed. However, the optical
uniformity of droplet size in emulsions obtained was observed by micrograph of the SPI-C/T emulsion (Fig. 3A) showed the formation
optical microscopy (Fig. 3A). The intense mechanical forces of coalesced T droplets with diameters from 5 to 10 mm. The
applied to the pre-emulsion during high-pressure homogenization improved hydrophilic character of SPI chains could reduce their
resulted in partial unfolding of proteinic chains and oil droplet affinity with hydrophobic T resulting in less efficient protein
dispersion. The emulsion was stabilized by an adsorbed layer of adsorption on the oil droplet surface and thus reduced protection
surface-active proteins, forming a protective barrier around the against coalescence and emulsion stability.
dispersed oil droplets. This protein layer provided immediate and Volume droplet size distributions of oil-in-water emulsions are
effective protection of the fine droplets against coalescence and shown in Fig. 3B. All the emulsions had droplet size distributions
gave a good stability to the emulsion. The decrease in droplet with explicit bimodal behavior. A second minor population with a
diameter observed for SPI-A/T emulsion compared to SPI/T emul- size of about 5  2 mm could be attributed to some coalesced
sion, could be due to the increased surface activity of SPI after droplets. The droplet size dispersion of SPI/T was narrower than
acylation, and thus better oil droplet dispersion. The same that of SPI-A/T, which confirmed that the droplet size distribution
behavior was observed for sunflower protein/a-tocopherol emul- was more uniform in the case of native protein based emulsion. It is
sion in our previous study (Nesterenko, Alric, Violleau, Silvestre, & important to note that this population of coalesced droplets was
Durrieu, 2013). larger in the SPI-C/T emulsion, which confirmed the decrease in
On the other hand, grafting of cationic quaternary ammonium emulsion stability.
groups to native SPI also involved a decrease in viscosity for both T
and AA liquid preparations. After cationization, proteinic matrix 3.2.2. Effect of SPI modification on the microencapsulation process
chains became more hydrophilic (Kiick-Fischer & Tirrell, 1998; A significant difference in spray-drying yield for emulsions with
Zohuriaan-Mehr et al., 2009). This increase in hydrophilic T (61e68%) compared to preparations with AA (77e87%) was

Fig. 3. (A) Optical micrographs (scale bar e 10 mm) and (B) droplet size distributions of T based emulsions with native (SPI/T), acylated (SPI-A/T) and cationized (SPI-C/T) soy
protein.
 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179 177

observed. This could be attributed to the hydrophobic and viscous microparticles was significantly influenced by the nature of feed
character of a-tocopherol that involved a higher microparticle liquid ingredients. In our case, the proteins were the most surface-
accumulation inside the drying chamber. In addition, an increase in active components in the liquid preparation and they would
microparticle size for spray-dried T based emulsions (7.6e9.3 mm) adsorb to the aireliquid interface of the drying droplets. This pref-
compared to AA based solutions (4.8e5.2 mm), was also due to a erential adsorption combined with protein film forming properties,
microparticle agglomeration effect induced by the presence of was responsible for the formation of a smooth protein skin on the
surface oil. Nevertheless, these results demonstrated that native microparticle surface (Jayasundera, Adhikari, Aldred, & Ghandi,
soy proteins could efficiently encapsulate hydrophobic (T) and 2009). SPI acylation increased hydrophobic and surface-active
hydrophilic (AA) core material with retention efficiencies of 79.7% character of proteinic chains. This modification was successful
and 91.8% respectively. Higher retention efficiency of AA compared with regard to T encapsulation, because of hydrophobic interactions
to T could be attributed to the hydrophilic character of SPI and its occurring between wall material and active core (Nesterenko et al.,
better affinity to the hydrophilic core. 2012; Nesterenko, Alric, Violleau, et al., 2013). After SPI acylation,
A wide range of biopolymers has been reported in the literature the hydrophobic microdomaines were formed in water because of
as efficient wall materials for ascorbic acid encapsulation by spray- hydrophobic association between fatty acid moities (Lee, Jo, Kwon,
drying (retention efficiency varying from 85% to 101%). These Kim Y.H., & Jeong, 1998). Thus, the formation of protein film on the
include: maltodextrin and cashew tree gum (Moreira, Azeredo, aireliquid interface was more effective, the exposure of oil to the
Medeiros, Brito, & Souza, 2010), gum Arabic and rice starch microparticle surface was minimized, and T was better protected.
(Trindade & Grosso, 2000), pea protein and carboxymethylcellulose Conversely, acylated SPI showed much less effectiveness in micro-
(Pierucci, Andrade, Baptista, Volpato, & Rocha-Leao, 2006). The encapsulation of a hydrophilic compound (AA), because of the
microencapsulation of a-tocopherol by spray-drying has been reduced affinity between core and wall material.
studied with different natural wall materials such as maltodextrin In contrast to the positive effect of SPI acylation on T retention
and gum Arabic (Faria et al., 2010), pea protein and carboxymeth- efficiency, the attachment of polar trimethylammonium functions
ylcellulose (Pierucci, Andrade, Farina, Pedrosa, & Rocha-Leao, to SPI by cationization resulted in the decrease of RE values for this
2007), and showed retention efficiency ranging from 73% to 87%. hydrophobic substance. This was due to the enhanced hydrophilic
Compared to literature values, the retention efficiencies obtained in character of cationized SPI, which led to lower emulsion stability.
this study are highly satisfactory. More oil could be expected at the microparticle surface and the loss
The results reported in Table 1 demonstrated that SPI acylation of active core during spray-drying was increased. On the other
enhanced the RE of the hydrophobic active substance (T) from 79.7% hand, the enhanced spray-drying yield for SPI-C/AA preparation
to 94.8% but reduced the RE of the hydrophilic active substance (AA) was observed because of better affinity between active core and
from 91.8% to 57.3%. During drying, the surface composition of modified wall material.

Fig. 4. (A) Scanning electron micrographs (external and internal structures, scale bar e 2 mm) and (B) particle size distributions of SPI-A/T and SPI/AA microparticles.
178 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179

3.2.3. Effect of SPI modification on microparticle size and Church, F. C., Swaisgood, H. E., Porter, D. H., & Catignani, G. L. (1983). Spectropho-
tometric assay using o-phthaldialdehyde for determination of proteolysis in
morphology
milk and isolated milk proteins. Journal of Dairy Science, 66, 1219e1227.
As can be clearly seen in Fig. 4A, the morphology of spray-dried Derkatch, S. R., Levachov, S. M., Kuhkushkina, A. N., Novosyolova, N. V.,
microparticles was characterized by spherical shape with di- Kharlov, A. E., & Matveenko, V. N. (2007). Rheological properties of concen-
ameters ranging from 1 to 10 mm. The surface of SPI microparticles trated emulsions stabilized by globular protein in the presence of nonionic
surfactant. Colloids and Surfaces, 298, 225e234.
was smooth and compact without presence of fissures. For the SPI- Dubey, R., Shami, T. C., & Bhasker, R. K. U. (2009). Microencapsulation technology
A/T sample, the inner structure of particles showed the presence of and application. Defence Science Journal, 59(1), 82e95.
small pores indicating that oil droplets were well distributed in the Faria, A. F., Mignone, R. A., Montenegro, M. A., Mercadante, A. Z., & Borsarelli, C. D.
(2010). Characterization and singlet oxygen quenching capacity of spray-dried
protein matrix. In the case of AA based microparticles, the disper- microcapsules of edible biopolymers containing antioxidant molecules. Jour-
sion of active core crystal powder in the protein matrix, involved nal of Agricultural and Food Chemistry, 58, 8004e8011.
the formation of a dense non-porous wall. Favaro-Trindade, C. S., Santana, A. S., Monterrey-Quintero, E. S., Trindade, M. A., &
Netto, F. M. (2010). The use of spray drying technology to reduce bitter taste of
The microparticle size distributions displayed in Fig. 4B and casein hydrolysate. Food Hydrocolloids, 24, 336e340.
results presented in Table 1, show that AA based microparticles Ga1uszkaa, A., Migaszewskia, Z., & Namiesnik, J. (2013). The 12 principles of green
(5.2 mm for SPI/AA) had lower average size compared to T based analytical chemistry and the SIGNIFICANCE mnemonic of green analytical
practices. Trends in Analytical Chemistry, 50, 78e84.
microparticles (9.3 mm for SPI/T). This variation in the particle size Goodno, C. C., Swaisgood, H. E., & Catignani, G. L. (1981). A fluorimetric assay for
was most likely due to the agglomeration of T based microparticles available lysine in proteins. Analytical Biochemistry, 115, 203e211.
because of the presence of surface oil. Moreover, the difference in Gouin, S. (2004). Microencapsulation: industrial appraisal of existing technologies
and trends. Trends in Food Science and Technology, 15, 330e347.
morphology of spray-dried droplets (homogeneous solution SPI/AA
Hua, Y., Cui, S. W., Wang, Q., Mine, Y., & Poysa, V. (2005). Heat induced gelling
or heterogeneous emulsion SPI/T) and in density of active core properties of soy protein isolates prepared from different defatted soybean
materials affected the size of microparticles produced. As reported flours. Food Research International, 38, 377e385.
in the literature, the expected size of microparticles with vegetable Jayasundera, M., Adhikari, B., Aldred, P., & Ghandi, A. (2009). Surface modification of
spray dried food and emulsion powders with surface-active proteins: a review.
protein produced by spray-drying, ranged from 2e3 mm (Pierucci Journal of Food Engineering, 93, 266e277.
et al., 2006; Pierucci et al., 2007) to 9e18 mm (Favaro-Trindade Kavaliauskaite, R., Klimaviciute, R., & Zemaitaitis, A. (2008). Interaction of starch
et al., 2010; Ortiz et al., 2009), which was comparable and within with glycidyltrimethylammonium chloride and properties of derived products.
Carbohydrate Polymers, 73, 665e675.
the particle size range obtained in this work. Kiick-Fischer, K. L., & Tirrell, D. A. (1998). Controlling absorbency in gelatin net-
works: preparation and characterization of alkylated, crosslinked gelatin.
Journal of Applied Polymer Science, 68, 281e292.
4. Conclusions Kim, Y. D., Morr, C. V., & Schenz, T. W. (1996). Microencapsulation properties of gum
arabic and several food proteins: spray-dried orange oil emulsion particles.
Journal of Agricultural and Food Chemistry, 44, 1314e1320.
This paper has demonstrated that some suitable modifications Lee, K. Y., Jo, W. H., Kwon, I. C., Kim, Y. H., & Jeong, S. Y. (1998). Structural deter-
of proteinic matrix material can increase the microencapsulation mination and interior polarity of self-aggregates prepared from deoxycholic
efficiency of different active compounds. For this, soy protein acid-modified chitosan in water. Macromolecules, 31, 378e383.
Matemu, A. O., Kayahara, H., Murasawa, H., Katayama, S., & Nakamura, S. (2011).
isolate (SPI) was modified by acylation and cationization and used Improved emulsifying properties of soy proteins by acylation with saturated
as wall material for a-tocopherol (T) and ascorbic acid (AA) fatty acids. Food Chemistry, 124, 596e602.
microencapsulation using the spray-drying technique. Moreira, G. E. G., Azeredo, H. M. C., Medeiros, M. F. D., Brito, E. S., & Souza, A. C. R.
(2010). Ascorbic acid and anthocyanin retention during spray drying of acerola
The attachment of a polar trimethylammonium function to SPI
pomace extract. Journal of Food Processing and Preservation, 34, 915e925.
by cationization affected its hydrophilic properties, solubility and Nesterenko, A., Alric, I., Silvestre, F., & Durrieu, V. (2012). Influence of soy protein’s
isoelectric point. Protein cationization resulted in reduced T structural modifications on their microencapsulation properties: a-tocopherol
microparticles preparation. Food Research International, 48, 387e396.
retention and in enhanced spray-drying yield for AA based prep-
Nesterenko, A., Alric, I., Silvestre, F., & Durrieu, V. (2013). Vegetable proteins in
arations. SPI acylation led to lower emulsion droplet size and microencapsulation: a review of recent interventions and their effectiveness.
viscosity because of enhanced surface-active properties. Due to Industrial Crops and Products, 42, 469e479.
the better affinity between hydrophobic core material and acyl- Nesterenko, A., Alric, I., Violleau, F., Silvestre, F., & Durrieu, V. (2013). New way of
biomaterials valorization: use of sunflower protein for a-tocopherol microen-
ated SPI, the retention efficiency was enhanced from 79.7% to capsulation. Food Research International, 53, 115e124.
94.8%. Microparticles obtained had spherical shape and an internal Nunes, M. C., Batista, P., Raymundo, A., Alves, M. M., & Sousa, I. (2003). Vegetable
morphology depending on the nature of the active core (porous proteins and milk puddings. Colloids and Surfaces, 31, 21e29.
Ortiz, S. E. M., Mauri, A., Monterrey-Quintero, E. S., & Trindade, M. A. (2009). Pro-
for liquid core and dense for crystal core). Thus, native and duction and properties of casein hydrolysate microencapsulated by spray dry-
modified SPI appear to be promising carriers for different delivery ing with soybean protein isolate. Food Science and Technology, 42, 919e923.
systems, such as nutraceutical, cosmetic or pharmaceutical Packer, J. E., Slater, T. F., & Willson, R. L. (1979). Direct observation of a free radical
interaction between vitamin E and vitamin C. Nature, 278, 737e738.
products. Pierucci, A. P. T. R., Andrade, L. R., Baptista, E. B., Volpato, N. M., & Rocha-
Leao, M. H. M. (2006). New microencapsulation system for ascorbic acid using
pea protein concentrate as coat protector. Journal of Microencapsulation, 23,
References 654e662.
Pierucci, A. P. T. R., Andrade, L. R., Farina, M., Pedrosa, C., & Rocha-Leao, M. H. M.
AOAC. (2007). Iodine in iodized salts. Method, 925.56 (18th ed.). Official Method of (2007). Comparison of a-tocopherol microparticles produced with different
Analysis. wall materials: pea protein a new interesting alternative. Journal of Microen-
Augustin, M. A., & Hemar, Y. (2009). Nano- and micro-structured assemblies for capsulation, 24, 201e213.
encapsulation of food ingredients. Chemical Society Reviews, 38, 902e912. Rampon, V., Riaublanc, A., Anton, M., Genot, C., & McClements, D. J. (2003). Evidence
Augustin, M. A., Sanguansri, L., & Bode, O. (2006). Maillard reaction products as that homogenization of BSA-stabilized hexadecane-in-water emulsion induces
encapsulants for fish oil powders. Journal of Food Science, 71(2), 25e32. structure modification of the nonadsorbed protein. Journal of Agricultural and
Baldaro, E., Pelizzari, R., Tenconi, M., & Li Bassi, G. (2008). Procedure for the prepa- Food Chemistry, 51, 5900e5905.
ration of puridied cationic guar. Patent N
WO2008058769. Rascon, M. P., Beristain, C. I., Garcie, H. S., & Salgado, M. A. (2011). Carotenoid
Caillard, R., Remondetto, G. E., & Subirade, M. (2009). Physicochemical properties retention and storage stability of spray-dried paprika oleoresin using gum
and microstructure of soy protein hydrogels co-induced by Maillard type cross- arabic and soy protein isolate as wall materials. Food Science and Technology, 44,
linking and salts. Food Research International, 42, 98e106. 549e557.
Channasanon, S., Graisuwan, W., Kiatkamjornwong, S., & Hoven, V. P. (2007). Rondel, C., Alric, I., Mouloungui, Z., Blanco, J.-F., & Silvestre, F. (2009). Synthesis and
Alternating bioactivity of multilayer thin films assembled from charged de- properties of lipoamino acid-fatty acid mixtures: influence of the amphiphilic
rivatives of chitosan. Journal of Colloid and Interface Science, 316, 331e343. structure. Journal of Surfactants and Detergents, 12, 269e275.
Charve, J., & Reineccius, G. A. (2009). Encapsulation performance of proteins and Rusli, J. K., Sanguansri, L., & Augustin, M. A. (2006). Stabilization of oils by
traditional materials for spray dried flavors. Journal of Agricultural and Food microencapsulation with heated protein-glucose syrup mixtures. JAOCS, 83,
Chemistry, 57, 2486e2492. 965e971.
 A. Nesterenko et al. / Food Hydrocolloids 38 (2014) 172e179 179

Tang, C. H., & Li, X. R. (2013). Microencapsulation properties of soy protein isolate Wong, P. Y. Y., Nakamura, S., & Kitts, D. D. (2006). Functional and biological activities
and storage stability of the correspondingly spray-dried emulsions. Food of casein glycomacropeptide as influenced by lipophilization with medium and
Research International, 52, 419e428. long chain fatty acid. Food Chemistry, 97, 310e317.
Trindade, M. A., & Grosso, C. R. F. (2000). The stability of ascorbic acid micro- Yu, C., Wang, W., Yao, H., & Liu, H. (2007). Preparation of phospholipid microcap-
encapsulated in granules of rice starch and in gum arabic. Journal of Microen- sules by spray drying. Drying Technology, 25, 695e702.
capsulation, 17, 169e176. Zohuriaan-Mehr, M. J., Pourjavadi, A., Salimi, H., & Kurdtabar, M. (2009). Protein-
Wang, J., Guo, J., Yue, T., Yuan, Y., Liu, X., & Kennedy, J. F. (2012). Cationization of and homo poly(amino acid)-based hydrogels with super-swelling properties.
Ganoderma lucidum polysaccharides in concentrated alkaline solutions as gene Polymers for Advanced Technologies, 20, 655e671.
carriers. Carbohydrate Polymers, 88, 966e972.