International Journal of Biological Macromolecules 120 (2018) 237–244

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

journal homepage: http://www.elsevier.com/locate/ijbiomac

Potential of taro starch spherical aggregates as wall material for spray

drying microencapsulation: Functional, physical and thermal properties
J. Hoyos-Leyva a, L.A. Bello-Pérez a,⁎, E. Agama-Acevedo a, J. Alvarez-Ramirez b
a
Centro de Desarrollo de Productos Bióticos, Instituto Politécnico Nacional, Km. 8.5 Carr. Yautepec-Jojutla, Colonia San Isidro, Apartado Postal 24, Yautepec, Morelos 62731, Mexico
b
Departamento de Ingeniería de Procesos e Hidráulica, Universidad Autónoma Metropolitana-Iztapalapa, Iztapalapa, Ciudad de México 09340, Mexico

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

Article history: Taro starch spherical aggregates have been considered as wall material for the microencapsulation of bioactive
Received 17 May 2018 compounds. The distribution of particle size, morphology, stability, glass transition temperature, ζ-potential,
Received in revised form 2 July 2018 physical properties and flowability and compression indexes, and functional properties (water, oil and dye reten-
Accepted 19 August 2018
tion capacity) were measured. The average diameter was 17.5 ± 0.3 μm. Aggregates formed by a relatively high
Available online 21 August 2018
number of starch granules were observed. The residual protein in the aggregates was in the periphery of the
Keywords:
starch granules, favoring the aggregation and the stability under aqueous stirring conditions. The ζ-potential
Starch spherical aggregates was −21.8 ± 0.3 mV. The value of the glass transition temperature was in the range from 176.8 to 75.4 °C, with
Glass transition 5% and excess water, respectively. The Hausner ratio and Carr index were 1.06 and 14.7, respectively. The oil
ζ-Potential holding capacity was 1.2 ± 0.01 g of oil·g−1 of dry solid. Overall, the functional, physical and thermal properties
of the spherical aggregates of taro starch granules showed that this material offers good potential for the micro-
encapsulation of bioactive compounds.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction
Starch spherical aggregates with small granule size (diameter b10
μm) are viable structures for the micro-encapsulation of bioactive com-
pounds [1]. The formation of spherical aggregates requires stability of
starch granules and the action of a bonding agent. The mechanism in-
volved in the formation of spherical aggregates is not clear at all, al-
though some results in the literature pointed out to the formation of
complexes between positively charged proteins located on the granule
surface and negatively charged polysaccharides. In an interesting
study, Debet and Gidley [2] showed that proteins play a central role in
the integrity of insoluble remnants after gelatinization of starch disper-
sions. In fact, proteins play the role of a bonding agent for maintaining
the cohesion of starch granules. It can be postulated that similar mech-
anisms could be involved in the formation of spherical aggregates. How-
ever, in many cases the fraction of proteins does not suffice to stabilize
the structure of spherical aggregates, such that an external bonding
agent (e.g., gum Arabic, carboxymethyl cellulose and carrageenan)
should be added. Taro starch is an interesting case where the fraction
and possibly the nature of proteins lead to the formation and stabiliza-
tion of spherical aggregates. Gonzalez-Soto et al. [3] reported that taro
starch has the ability of forming spherical aggregates without the addi-
tion of bonding agents. In fact, isolated taro starch contains a relatively
⁎ Corresponding author.
E-mail address: labellop@ipn.mx (L.A. Bello-Pérez).
high content of protein (5–7%), which suffices for bonding starch gran-
ules within a spherical configuration. The characterization of taro starch
spherical aggregates focused on the study of the changes within the in-
ternal structure of the starch granules influenced by the spray drying
process and the effect of the binding agents addition at different ratios
[2–4]. However, other characteristics such as the glass transition tem-
perature, the stability under water and stirring conditions, the ζ-
potential, physical and functional properties remain scarcely explored.
These characteristics provide valuable information for assessing the po-
tential of taro starch spherical aggregates to obtain a stable wall material
and for the protection of a core material.
The characteristics of the wall material, such as chemical structure,
thickness, pore size and surface integrity, play a determining role in
the transport and release properties of the core material and on the sta-
bility of the microcapsules. It is well-known that the diffusion through a
material is facilitated by the rubbery state, while in the vitreous state the
diffusion is relatively slow [6]. The plasticizing effect on the wall mate-
rial induced by the glass transition can cause the ordering or rupture
of the polymer structure, resulting in undesirable changes of the charac-
teristics of a microencapsulant matrix [7]. Commonly, the glass transi-
tion temperature is estimated by calorimetric techniques. In the
modulated differential scanning calorimetry (MDSC), the sample un-
dergoes a sinusoidal oscillation of the temperature superimposed on a
heating ramp of the conventional calorimetry, so that the total heat
flow can be separated in its heat capacity (reversible heat flow) and ki-
netic components (irreversible heat flow) [8,9]. On the other hand, the

https://doi.org/10.1016/j.ijbiomac.2018.08.093
0141-8130/© 2018 Elsevier B.V. All rights reserved.
238 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244
production of microcapsules involves in general high shear and elonga-
tion conditions with significant temperature gradients. Mechanical
forces and temperature gradients in food processing equipment could
induce substantial deformations of a microcapsule, leading to rupture
of the wall material and premature release of the core material [10]. Sta-
bility of microcapsules under stringent stirring conditions is an issue
that has been rarely addressed.
The objective of this study was to assess the characteristics of taro
starch spherical aggregates, such as the size and morphology, stability
under agitation in aqueous medium, ζ-potential, glass transition tem-
perature, bulk density, absolute density, particle density, apparent po-
rosity, flowability and compressibility indexes, and some functional
properties (water, oil and dye retention capacity). These characteristics
would provide valuable insights in the ability of taro starch spherical ag-
gregates for providing protection to bioactive compounds, as well as for
estimating optimal process conditions for the application of spherical
aggregates.
2. Materials and methods
2.1. Preparation of spherical aggregates
Taro flour was produced from collected corms of Colocassia esculenta
var. Esculenta harvested in a commercial crop at Tuxtepec, Mexico.
Starch isolation was carried out following the Whistler patent [11] pro-
cedure. The starch was filtered through 50, 100 and 325 US mesh. After
the final washed of starch, the solid concentration of the starch slurry
was adjusted to 30 g · 100 g−1 water. The starch dispersion was fed to
a Mini Spray Dryer (Model B-290, BÜCHI UK Ltd., Chadderton, UK)
equipped with a double flow atomizer. The drying conditions were
the previous reported by Gonzalez-Soto et al. [3], with inlet temperature
145 °C, outlet temperature 80 °C and flow rate of 7.6 g·min−1. The dried
powder was stored in a glass container to further analysis. Before spray
drying, a fraction of starch slurry was centrifuged at 10,800g for 15 min,
and the precipitated starch paste was transfer to a forced convection
oven (40 °C, 24 h) to obtain native taro starch conventional dried [12].
This sample was used for comparison purposes.
2.2. Preparation of taro starch spherical aggregates-low protein content
Conventionally dried native taro starch (3.0 g) was dispersed in
25 mL NaOH (0.2 g · 100 mL−1). The dispersion was magnetic stirred
for 2 h. After stirring, the dispersion was centrifuged (5000g for
5 min), the supernatant was discarded and the precipitated was washed
with water until achieved pH ~7. An aqueous dispersion (30 g · 100 g−1)
was prepared with the precipitated starch and was spray dried under
some conditions used to taro starch spherical aggregates preparation.
2.3. Water activity, moisture, protein and starch content
Moisture and protein content were tested following the AACC
methods 44-15.02 and 45-13.01, respectively [13]. Total starch was de-
termined by means of the total starch kit procedure from Megazyme In-
ternational Ireland Ltd., using the Item C for samples containing
resistant starch. Water activity was tested in a paw kit instrument
(Decagon Devices, Inc., Pullman, Washington, US).
2.4. Particle size distribution and morphology
The size distribution of spherical aggregates was measured using a
Mastersizer 3000 instrument (Malvern Instruments Ltd., Malvern,
Worcestershire, UK) equipped with a Hydro EV accessory. The sample
was diluted in water until 12–15% laser obscuration. The sample mea-
sure time was 15 s without sonication. The sample was analyzed by trip-
licate. The morphology of the starch spherical aggregates was observed
with a JSM 5600LV microscopy (JEOL, Tokyo, JP), the resolution was
5 nm in low vacuum (LV) mode. The instrument conditions were
20 kV electron acceleration voltage, 12–20 Pa of pressure, and the im-
ages were obtained from backscattering electron signal. Additionally,
taro starch spherical aggregates were observed in a confocal laser scan-
ning microscope (LSM 710 NLO, Carl Zeiss Microscopy GmbH, Jena, DE)
in stack mode, filters 417–729, laser at 405 nm (50.0%), 561 nm (6.5%),
468 (7.5%) and 633 nm (8.0%). Previous scanner was done to test the au-
tofluorescence region of protein fraction.
2.5. Stability under stirring conditions
Stirring operations are typically observed in the food industry. The
encapsulated biocompounds added to a food product should resist the
conditions of mixing or homogenizing. The stability of spherical aggre-
gates was tested following the Zhao and Whistler [1] procedure. Briefly,
a sample of spherical aggregates was dispersed in water (5 g · 100 g−1)
at 25 °C by stirring at 100 rpm for 24 h. The shear-rate conditions were
about 1.8 s−1 with Reynolds number for cylindrical tanks of the order of
12. An axial impeller was used for inducing turbulent flows with both
elongational and rotational shear effects. Aliquots were takes each
hour and analyzed in the Mastersizer 3000 instrument under the
same conditions used for particle size distribution analysis.
2.6. ζ-Potential
The surface charge was measured as ζ-potential. The ζ-potentials of
spherical aggregates and native taro starch (oven dried) were assessed
in a Zetasizer Nano Z instrument (Malvern Instruments Ltd., Malvern,
Worcestershire, UK). The measurement was replicated three times.
2.7. Glass transition temperature
The evaluation of the effect of relative humidity (RH) on the glass
transition temperature was evaluated. To this end, taro starch spherical
aggregates were exposed to a different relative humidity conditions in
desiccators containing saturated salt solutions. The values of the RH
were 13% (LiCl), 26% (MgCl2), 46% (K2CO3), 57% (NaBr) and 72%
(NaCl) at 26 °C. The samples were stored under those RH conditions
for 2 weeks. The moisture was assessed to assure the water content in
the sample. The glass transition temperature of taro starch spherical ag-
gregates in water excess and RH conditioned was assessed by modu-
lated temperature differential scanning calorimetry (MTDSC) using a
DSC instrument (TA Instrument, Q20, New Castle, NJ, US) in modulated
temperature setting. For analysis under water excess, the sample
(2.2 mg d.b) was weighted in aluminum pan with 80% (w/w) of deion-
ized water. The temperature profile consisted in two heating from 10 to
110 °C, 2 °C·min−1 heating rate, modulation conditions amplitude 0.5 °C
and frequency of 60 s. For RH conditioned, 5 mg of sample was weighted
and the heating rate of MTDSC was 5 °C·min−1. The samples were run by
triplicated. The analysis of the glass temperature (Tg) was made directly
from the instrument software. The reversible heat flow was used to the
analysis.
2.8. Density, apparent porosity, flowability and compressibility index
The physical properties were measured for native taro starch, taro
starch spherical aggregates and native maize starch. The latter sample
was used as reference material since the physical properties of maize
starch has been widely studied. The bulk density (db) was determined
from 3 g of taro starch spherical aggregates weighted in a 25 mL volu-
metric graduated cylinder. The occupied volume by the sample was reg-
istered. The particle density (dp) was tested with the previous weighted
sample, which was mechanically taped until constant volume. The
 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244 239

apparent porosity (ε) was calculated with the following equation [14]: comparative test (significance level 0.05). The statistical analysis was
carried out using JMP v. 10.0 software (SAS Institute Inc., Cary, NC, US).
db
ε ¼ 1− ð1Þ
dp 3. Results and discussion

The absolute density (da) was calculated by the relation between
the displaced volume of a known sample weight in ethanol (99.0%) in
a volumetric graduated cylinder. Flowability index was measured as
Hausner ratio (HR), which indicate the flowability of a granular or pow-
der solid [15,16]. The HR was determined following the methodology
proposed by Akhavan Mahdavi et al. [17], where HR is the ratio between
onset volume (V0) and final volume (Vf) registered to bulk and particle
density measures. The compressibility was quantified as Carr's index
(CI) with the following expression:
V 0 −V f
CI ¼  100
Vf
2.9. Functional properties of spherical aggregates
Water retention capacity (WRC), and oil and colorant absorption ca-
pacity were determined following the methodology reported by Guo
et al. [18]. The colorant used for the test was a solution of blue methy-
lene (30 mg·L−1). A maximum wavelength was tested to quantify the
colorant absorbed by the taro starch spherical aggregates.
2.10. Statistical analysis
All analyses were carried out by triplicate. Mean and standard error
(SE) were reported. Analysis of variance was used to stablish statistical
differences (p b 0.05) and Tukey test was employed as mean
3.1. Chemical composition of starch spherical aggregates
The moisture content and water activity of taro starch spherical ag-
gregates were 2.3 g · 100 g−1 and 0.33, respectively. The relatively low
value of moisture content is related to the high efficiency of the drying
process and the aggregation of starch granules that limit water reten-
tion, resulting in a well dried powder.
The protein content (3.7 ± 0.03 g · 100 g−1) was lower than the
value reported by Gonzalez-Soto et al. [3] (4.5 g · 100 g−1). However,
Fig. 1 suggested that this amount of protein suffices for the aggregation
of the starch granules due to the hydrophilic character of protein and
ð2Þ
starch chains. Results in previous studies showed that a small amount
of indigenous protein (0.34 g · 100 g−1) was sufficient to produce spher-
ical aggregates of amaranth starch [1]. Differences in the protein content
may be due to the isolation process and differences in crop agroclimatic
conditions. It has been reported that the formation of starch spherical
aggregates can be achieved with the addition of bonding agents
(e.g., gum Arabic, carboxymethyl cellulose and carrageenan) with per-
centages ranging in 0.1 to 2%, although without the incorporation of
proteins [1,4,5,19]. The protein content of the spray dried taro starch
spherical aggregates was relatively low (about 0.5 ± 0.01 g · 100 g−1).
The starch content (83.3 ± 0.8 g · 100 g−1 d.b.) is related to the
starch purity and is a measure of starch isolation efficiency. Starch purity
has been reported in the range from 75 to 95% [3,20], the differences in
this value may be due to taro variety, harvest stage and starch granules
size, since in the isolation process the starch slurry was filtered through
325 US mesh.

a b

c d

Fig. 1. Spherical aggregates from spray dried taro starch observed in E-SEM: a) 2000×; b) 10,000×; and light microscopy (c and d, 100× objective lens).
240 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244
3.2. Morphology and particle size distribution of spherical aggregates
The morphology of microcapsules is illustrated in Fig. 1 for 2000×
(a) and 10,000× (b). The spherical aggregates exhibited a diversity of
sizes, and in some cases the shape was slightly irregular due maybe to
the polyhedral shape of individual taro starch granules [12]. The spher-
ical aggregates showed cavities that can be exploited for loading differ-
ent core materials. Previous reports have shown the formation of
spherical aggregates with small granular starches isolated from diverse
botanical sources (e.g., amaranth, small wheat starch granules, and
rice). Zhao and Whistler [1] reported that the addition of binding agents
(0.1–1.0%) was necessary for the formation of starch spherical aggre-
gates of diverse botanical sources. However, Beirão-Da-Costa et al. [4]
reported the formation of starch spherical aggregates with rice starch,
when the protein content was not higher than 1%. Also, Gonzalez-Soto
et al. [3] reported the formation of spherical aggregates with taro starch
with relatively high protein content of about 4.5%. The light microscopy
images (Fig. 1.c and d) revealed spherical aggregates with distinct size,
formed by a relative widely number of starch granules. On the other
hand, confocal laser scanning microscopy (Fig. 2) revealed the protein
localization in the periphery of taro starch granules. The protein was
identified by the blue color in the autofluorescence image (Fig. 2.a and
c). This finding corroborated the hypothesis of Gonzalez-Soto et al. [3]
about the influence of residual protein on the spherical aggregates for-
mation. In fact, Fig. 2.c evidenced that protein acts as sticking material
facilitating the aggregation of starch granules.
The particle size distribution of the spherical aggregates exhibited a
monomodal distribution (Fig. 3). Some studies have reported that
a b
Fig. 2. Confocal laser scanning microscopy images of taro starch spherical aggregates: protein autofluorescence (a), light microscopy image (b) and a/b superposition (c).
double-flow atomizer produce uniform particle size in the spray drying
[21]. The particle size distribution is a relevant variable that could be
assessed for the encapsulation of compounds within starch spherical ag-
gregates, since differences in diameter could affect the stability of the
microcapsule, the potential volume to core material retention, and its
controlled release. In addition, particle size is a feature that can also in-
fluence consumer acceptance of a potential food product in which taro
starch spherical aggregates potentially will be used as carrier material
to biocompounds or microorganisms [22].
3.3. Spherical aggregates stability to stirring conditions
The stability of taro starch spherical aggregates measured as the par-
ticle size change in dispersion in water with constant agitation
(100 rpm) at 25 °C was summarized in Table 1. The spherical aggregates
retained their configuration for 1 h under stirring conditions. After 3 h,
reductions in the size distribution were observed, a trend that was ob-
served until 24 h of constant stirring. On the other hand, taro starch
spherical aggregates-low protein content (0.5%) showed low percent-
ages of volume to particles N10 μm when dispersed in water, which in-
dicates that the starch granules were easily dispersed in water. This
behavior suggested that the taro starch spherical aggregates were un-
stable under stirring conditions. However, the protein content (3.7 ±
0.03 g · 100 g−1) had a positive effect in aggregates stability, since aggre-
gates with low protein sample were easily dispersed in water. The larger
spherical aggregates (Dx (90)) were preserved up to 7 h under stirring
conditions in water (Table 1), which indicates that aggregates of larger
size, with a greater number of aggregate starch granules, are more
 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244 241

14 granule, such that aggregation is possible thanks to the sticky effect of

proteins under high-rate heating in the drying process.
12

10 3.5. Glass transition temperature

Volume (%)

8
6 of a well-defined endothermic heat flow linked to gelatinization is ap-
4 shown in Fig. 4.b. The value of Tg was reflected in the change of revers-
2
0 onset gelatinization temperature. Differences in the Tg value of polysac-
1 10 100
Particle Size (µm)
Fig. 3. Particle size distribution of native taro starch (dashed line) and taro starch spherical
aggregates (continuous line).
stable structures and retain their structure for a longer time as com-
pared with smaller aggregates (e.g., 17.5 μm).
3.4. ζ-Potential
The ζ-potential or electrokinetic potential is a feature of the surface
of particles. The measure of ζ-potential is a useful data to analyze possi-
ble starch interactions with other components or between two or more
granules. For instance, the ζ-potential is a measure of the electrostatic
potential and Van der Walls forces that facilitated those interactions.
The ζ-potential of the starch spherical aggregates was −21.8 ± 0.3 mV.
The negative value of ζ-potential has been reported for potato and
wheat starches [23], isolated starches from Dioscorea spp. [24] and na-
tive and modified rice starch (phosphorylation and cross-linking) [25].
The starch phosphorylation cause an increase in the negative value of
ζ-potential, which was attributed to the phosphate group [25]. The
taro starch spherical aggregates were disaggregated by sonication to
measure the ζ-potential of disintegrated taro starch granules, resulting
in ζ-potential values of −20.3 ± 0.3 mV. This indicates that the aggrega-
tion of the taro starch granules may be influenced by the surface charge,
which is reduced when they are disintegrated in water. Theoretically, it
has been reported that absolute values greater than ±30 mV represent
high stability of dispersed particles, given that they have greater interac-
tion with compounds of opposite charges. It has been postulated that
proteins and lipids play a central role in the integrity of native starch
granules and insoluble remnants (a.k.a., ghosts) after gelatinization
[2]. Positively charged proteins have been observed on the starch
The glass transition was estimated from reversible heat flow of the
modulated DSC analysis. Fig. 4.a illustrates the whole DSC pattern, in-
cluding both forward and backward (i.e., reversible paths). The absence
parent. The corresponding heat flow transition for estimation of Tg is
ible heat flow (delimited by the parallel red lines), and the value of Tg
was determined as the midpoint of the thermal transition. Xie et al.
[26] reported a Tg value close to 60 °C for waxy corn starch under excess
moisture (77%). In a native starch Tg is localized immediately before the
charides have been attributed to variations in the average molar mass
weight and in the case of starches, to the variation of the amylose:amy-
lopectin ratio [26,27]. The value of Tg can be interpreted from two
points of view. The first is focused on spray drying and the formation
of spherical aggregates, since the role of the glass transition in the for-
mation of agglomerates in food products has been reported [29].
Turchiuli et al. [30] reported that maltodextrins with different dextrose
equivalents (12 and 21 DE) showed a sticky state when the particles
were close to the atomizer outlet in the spray dryer chamber, while
maltodextrin with 12 DE showed this state throughout the drying
chamber. It should be noted that maltodextrin 21 DE had a lower Tg
(~100 °C) than the maltodextrin 12 DE (Tg~120 °C), both with maxi-
mum moisture content of 6 g · 100 g−1. Fig. 4 suggests that the starch
is a rubbery state under excess of water and during drying changes to
a glassy state. The second point of view on Tg is focused on the stability
of spherical starch aggregates for use in food products that involves hu-
midity and temperature conditions. To ensure physical stability of taro
starch spherical aggregates under excess water, the process tempera-
ture must be lower than Tg (75.4 °C).
The value of Tg was higher for low moisture content (low RH %) in
the sample (Fig. 4.c), a result that was in line with the Tg values reported
for the plasticizing effect of water on starch [31]. The highest value of Tg
(176.8 °C) was found with 5% RH and the lowest value (89 °C) was ob-
served with 72% RH (Fig. 4.c). These values correspond to worst-case
conditions, so they provide valuable insights on the stability of spherical
aggregates. In fact, the glass transition temperature is a key parameter
that can be tuned up for avoiding physical changes in the structure of
the encapsulating material. Data of Tg and humidity should be consid-
ered for product storage, since information on agglomeration avoiding
or formation of lumps in the material can be predicted. On the other
hand, the values of Tg indicated that the stability of taro starch spherical
aggregates is comparable to the stability exhibited by commonly used
wall material, such as amylopectin [32], cassava starch films [33] and
starches with different amylose:amylopectin ratio [28]. That is,

Table 1
Percentiles of size distribution of starch spherical aggregates under stirring conditions in water with two protein content (g · 100 g−1).

Time (h) Protein content

3.7 0.5

Dx (10) Dx (50) Dx (90) Dx (10) Dx (50) Dx (90)

0 1.12 ± 0.01 18.9 ± 0.1 37.4 ± 0.1 0.996 ± 0.11 4.15 ± 0.10 17.7 ± 0.1
1 1.14 ± 0.1 17.5 ± 0.3 36.2 ± 0.2 1.07 ± 0.01 3.72 ± 0.01 15.3 ± 0.02
3 1.08 ± 0.02 8.83 ± 0.05 31.3 ± 0.01 1.0 ± 0.01 3.49 ± 0.01 13.9 ± 0.05
5 1.06 ± 0.05 7.16 ± 0.4 30.1 ± 0.02 1.01 ± 0.01 3.53 ± 0.01 14.9 ± 0.01
7 1.06 ± 0.01 5.79 ± 0.02 31 ± 0.1 1.01 ± 0.01 3.24 ± 0.03 13.2 ± 0.1
24 1.01 ± 0.01 3.19 ± 0.1 19.6 ± 0.6 1.01 ± 0.01 3.04 ± 0.01 12.2 ± 0.02
Referencea 0.85 ± 0.1 1.43 ± 0.01 4.3 ± 0.04 0.85 ± 0.01 1.43 ± 0.1 4.3 ± 0.04

Dx(y): Particle diameter smaller than “y”. Values are means of three replicates ± SE.
a
Disaggregates taro starch granules under sonication (100%) previous particle size distribution analysis.
242 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244

0.4 weight of 13 × 108 g·mol−1 [36]. Statistical differences were found be-
Heat Flow (W g )

(a)
-1

0.3 tween the values of bulk and particle densities for the native taro starch
0.2
and spherical aggregates. However, in the absolute density measured by
the volume displaced by a known weight of the material, only marginal
0.1 differences were observed. This effect could be caused by disintegration
0.0 of starch granules, which result in values of absolute density close to na-
-0.1 tive starch.
The increase of the apparent porosity in the spherical aggregate of
-0.2
taro starch with respect to the native starch is an indication of the po-
0 50 100 150 200
rous structure of the starch spherical aggregates. The value of the poros-
ity ε for the taro starch spherical aggregates was about 30% higher than
Reversible Heat Flow

(b)
the value for native taro starch. This relatively low difference may be
due to the small size of the taro starch granules (Fig. 3), which have ir-
(W g )

regular and polyhedral shapes [12]. This gives rise to porous packing or
-1

inter-particle cavities during density measurements.

The Hausner ratio (HR) was 1.1 and 1.06 for native taro starch and
taro starch spherical aggregates, respectively (Table 2). These values in-
dicate free flowability of starchy powders [15]. The free flowability is a
69 72 75 78 81 84 desired feature in materials used as excipients in the manufacture of
o
Temperature ( C) tablets for pharmaceutical and cosmetic industries. The Carr index (CI)
180 is a value that indicates the compressibility of the material. Taro starch
(c)
160 spherical aggregates had a CI of 14.9, higher than native taro starch
-0.27 2 (10.0) and native corn starch (9.6) (Table 2). The value of 14.9 is an in-
140 y = 278.9x , R = 0.93
dicator that the product can compress and is useful for diverse applica-
Tg ( C)

tions, such as the formation of tablets in the pharmaceutical industry.

o

120
The higher value of the spherical aggregates with respect to the native
100 taro starch could be due to structures formed by the aggregation of
80 the starch granules, which can be compressed. In this regard, the use
0 10 20 30 40 50 60 70 80 of spherical aggregates of native and modified rice starch as excipients
Relative Humidity (%) for the pharmaceutical industry has been reported [37–39], which
may open a field of research for the taro starch spherical aggregates.
Fig. 4. (a) Example of heat flow pattern including forward and backward (reversible)
paths, (b) observation of the glass transition temperature in the reversible heat flow 3.7. Taro starch spherical aggregates functional properties
path, (c) glass transition temperature as function of the relative humidity.

To evaluate the effect of the aggregation of the taro starch granules

spherical aggregates offered sufficient stability to be considered for mi-
croencapsulation of bioactive compounds.
3.6. Bulk, particle and absolute density, apparent porosity, flowability and
compressibility indexes
The bulk, particle and absolute density, as well as the apparent po-
rosity and the flowability and compressibility indexes of taro starch ag-
gregates are summarized in Table 2. The bulk density is a property
related to the size, shape and surface properties of the particles [34].
The bulk density of the native taro starch was higher than taro starch
spherical aggregates produced, an effect that could be resulting from
the porous structure produced by the aggregation of the starch granules
during spray drying. In turn, this leads to powders with lower bulk den-
sity. It has been reported that the bulk density is higher when the parti-
cles are smaller and with regular shapes, which produces a more
uniform arrangement with fewer inter-particle empty spaces [34,35].
In fact, the amylopectin fraction in taro starch presented a molecular
Table 2
Physical properties, Hausner ratio and Carr index of native taro starch (NTS), taro starch
spherical aggregates (TSSA) and native maize starch (NMS)⁎.
on the retention of water, oil and a dye, the native taro starch was com-
pared with the spherical aggregates (Table 3). The spherical aggregates
of taro starch exhibited higher values of WRC and OAC than the native
taro starch counterpart. However, the native starch shown the higher
value of the DAC (23 mg·g−1 dry solid) than the spherical aggregates.
The OAC results could be related to the macroporous structure of the
spherical aggregates, with the ability of retaining more oil. These values
found for the water retention capacity were higher than those reported
by Wootton and Bamunuarachchi [40] for native starches isolated from
different botanical sources (0.26–0.38 g H2O·g−1 dry starch), native corn
starch (0.40 g H2O·g−1 dry starch) and enzymatically modified and
cross-linked starches (0.51–0.65 g H2O·g−1 dry starch) [18]. Differences
in water retention have been related to the surface area and the porosity
of the material.
The oil absorption capacity (OAC) and DAC of taro starch spherical
aggregates were higher than those reported for porous starches pro-
duced by enzymatic modification, exhibiting values in the range
0.60–0.62 g oil·g−1 of dry starch and 18–20 mg of dye·g−1 of dry starch,
respectively [18]. The OAC of taro starch spherical aggregates was
higher than the OAC of porous starch used for microencapsulation of
olive oil [41]. This suggests that spherical taro aggregates have a larger

NTS TSSA NMS Table 3

Bulk density (g·cm−3) 0.50 ± 0.01a 0.37 ± 0.06b 0.43 ± 0.05c Water retention capacity (WRC), oil adsorption capacity (OAC) and dye adsorption capac-
Particle density (g·cm−3) 0.60 ± 0.01a 0.44 ± 0.01b 0.48 ± 0.01c ity (DAC) of native taro starch and taro starch spherical aggregates⁎.
Absolute density (g·cm−3) 1.02 ± 0.05a 1.10 ± 0.03b 1.10 ± 0.05b
Sample WRC (g·g−1) OAC (g·g−1) DAC (mg·g−1)
Apparent porosity 0.40 ± 0.01a 0.70 ± 0.01b 0.61 ± 0.04c
Hausner ratio 1.10 ± 0.01a 1.06 ± 0.01b 1.10 ± 0.05ab Native taro starch 1.4 ± 0.01 a
1.0 ± 0.01 a
23.2 ± 0.9a
Carr index 10.0 ± 0.11a 14.7 ± 0.01b 9.60 ± 0.42a Taro starch spherical aggregates 1.5 ± 0.01b 1.2 ± 0.01b 19.6 ± 1.0b
⁎ Values are the mean of three replicates ±SE. Different letter in row indicate significant ⁎ Values are the mean of three replicates ±SE. Different letter in column indicate sig-
statistical differences (p b 0.05). nificant statistical differences (p b 0.05).
 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244
pore volume than porous starches used in the food industry, with po-
tential for microencapsulation of diverse core materials, such as essen-
tial oils, lipophilic vitamins, etc. On the other hand, DAC was higher
for native taro starch than for the corresponding spherical aggregates
(Table 3). Dyes can be adsorbed by protein matrices via the formation
of complexes [42]. Under the postulate that protein acted as a sticky
agent, the available protein for dye adsorption is effectively reduced
by the formation of the spherical aggregates. As a consequence, the
DAC of native starch granules is superior to that exhibited by their
spherical aggregates.
4. Conclusions
The characteristics of the taro starch spherical aggregates focusing
on their application as wall material in microencapsulation purposes
were assessed in this study. The stability of spherical aggregates in
water under stirring conditions was low, which can be indicating that
these structures would not be suitable to encapsulate biocompounds
intended for addition in liquids (e.g., juices, drinks, etc.). However, po-
tential applications in viscous food and cosmetics or low moisture con-
tent products are viable. The glass transition temperature was higher
under low humidity, and the glass temperature decreased as the humid-
ity increased. These properties should be considered both for the use of
this material in food products that require humidity and temperature
conditions, and to establish storage conditions that do not affect the
physical structure of the material. From the evaluated physical proper-
ties (apparent, particle and absolute density, as well as the flowability
and compressibility indexes), it can be concluded that the spherical ag-
gregates could be used as wall material, as excipient to produce tablets.
From the glass transition temperature and the ζ-potential, it can be hy-
pothesized that these two properties partially explain the aggregation
of taro starch granules. The effect of the protein on the aggregation
was clarified for the water stability analysis, which was lower when
the starch contained less protein. In addition, confocal microscopy im-
ages showed protein location in the periphery of the aggregated starch
granules.
Acknowledgments
We appreciate the economic support from SIP-IPN, EDI-IPN
CONACYT-México and COFAA-IPN. JDHL also acknowledges the scholar-
ship from CONACYT-Mexico.
Conflict of interest
The authors declare no conflict of interest.
References
[1] J. Zhao, R.L. Whistler, Spherical aggregates of starch granules as flavor carriers, Food
Technol. 48 (1994) 104–105.
[2] M.R. Debet, M.J. Gidley, Why do gelatinized starch granules not dissolve completely?
Roles for amylose, protein, and lipid in granule “ghost” integrity, J. Agric. Food Chem.
55 (2007) 4752–4760, https://doi.org/10.1021/jf070004o.
[3] R.A. Gonzalez-Soto, B. de la Vega, F.J. García-Suarez, E. Agama-Acevedo, L.A. Bello-
Pérez, Preparation of spherical aggregates of taro starch granules, LWT Food Sci.
Technol. 44 (2011) 2064–2069, https://doi.org/10.1016/j.lwt.2011.06.018.
[4] S. Beirão-Da-Costa, C. Duarte, M. Moldão-Martins, M.L. Beirão-Da-Costa, Physical
characterization of rice starch spherical aggregates produced by spray-drying, J.
Food Eng. 104 (2011) 36–42, https://doi.org/10.1016/j.jfoodeng.2010.11.024.
[5] T.a. Tari, U.S. Annapure, R.S. Singhal, P.R. Kulkarni, Starch-based spherical aggre-
gates: screening of small granule sized starches for entrapment of a model
flavouring compound, vanillin, Carbohydr. Polym. 53 (2003) 45–51, https://doi.
org/10.1016/S0144-8617(02)00293-X.
[6] M.I. Ré, Microencapsulation by spray drying, Dry. Technol. 16 (1998) 1195–1236,
https://doi.org/10.1080/07373937.2015.1010209.
[7] C. Anandharamakrishnan, P. Ishwarya, Spray Drying Technique for Food Ingredient
Encapsulation, John Wiley and Sons, Ltd, 2015 https://doi.org/10.1002/
9781118863985 (First).
243
[8] A. Boller, C. Schick, B. Wunderlich, Modulated differential scanning calorimetry in
the glass transition region, Thermochim. Acta 266 (1995) 97–111, https://doi.org/
10.1016/0040-6031(95)02552-9.
[9] M. Reading, A. Luget, R. Wilson, Modulated differential scanning calorimetry,
Thermochim. Acta 238 (1994) 295–307, https://doi.org/10.1016/S0040-6031(94)
85215-4.
[10] L.M.C. Sagis, Microencapsualtion and Microsperes for Food Applications, Academic
Press, 2015.
[11] R.L. Whistler, US patent 5797985, 1998.
[12] J.D. Hoyos-Leyva, L.A. Bello-Pérez, H. Yee-Madeira, M.E. Rodriguez-Garcia, A.
Aguirre-Cruz, Characterization of the flour and starch of aroid cultivars grown in
Mexico, Starch-Starke 69 (2017) 1–8, https://doi.org/10.1002/star.201600370.
[13] A.A. of C.C. AACC, Approved Methods of the American Association of Cereal Chemist,
10th ed. AACC International, St Paul, MN, 2000.
[14] M.K. Krokida, V.T. Karathanos, Z.B. Maroulis, Effect of freeze-drying conditions on
shrinkage and porosity of dehydrated agricultural products, J. Food Eng. 35 (1998)
369–380, https://doi.org/10.1016/S0260-8774(98)00031-4.
[15] H. Hausner, Friction conditions in a mass of metal powder, Int. J. Powder Metall. 3
(1967) 713.
[16] G. Lumay, F. Boschini, K. Traina, S. Bontempi, J.C. Remy, R. Cloots, N. Vandewalle,
Measuring the flowing properties of powders and grains, Powder Technol. 224
(2012) 19–27, https://doi.org/10.1016/j.powtec.2012.02.015.
[17] S. Akhavan Mahdavi, S.M. Jafari, E. Assadpoor, D. Dehnad, Microencapsulation opti-
mization of natural anthocyanins with maltodextrin, gum Arabic and gelatin, Int. J.
Biol. Macromol. 85 (2016) 379–385, https://doi.org/10.1016/j.ijbiomac.2016.01.011.
[18] L. Guo, R. Liu, X. Li, Y. Sun, X. Du, The physical and adsorption properties of different
modified corn starches, Starch-Starke 67 (2015) 237–246, https://doi.org/10.1002/
star.201400200.
[19] M. Trindade, C.R.F. Grosso, The stability of ascorbic acid microencapsulated in gran-
ules of rice starch and in gum arabic, J. Microencapsul. 17 (2000) 169–176.
[20] E. Agama-Acevedo, F.J. Garcia-Suarez, F. Gutierrez-Meraz, M.M. Sanchez-Rivera, E.
San Martin, L.A. Bello-Pérez, Isolation and partial characterization of Mexican taro
(Colocasia esculenta L.) starch, Starch-Starke 63 (2011) 139–146, https://doi.org/
10.1002/star.201000113.
[21] I.C. Kemp, T. Hartwig, R. Herdman, P. Hamilton, A. Bisten, S. Bermingham, Spray dry-
ing with a two-fluid nozzle to produce fine particles: atomization, scale-up, and
modeling, Dry. Technol. 34 (2016) 1243–1252, https://doi.org/10.1080/07373937.
2015.1103748.
[22] K. Adhikari, Microencapsulated bifidobacterium longum in stirred yogurt, J. Food
Sci. 68 (2003) 275–280.
[23] R.A. Marsh, S.G. Waight, The effect of pH on the zeta potential of wheat and potato
starch, Starch-Starke 34 (1982) 149–152, https://doi.org/10.1002/star.
19820340502.
[24] G.K. Gbassi, C.N. Amin, P.G. Atheba, A. Trokourey, Granulometric structure, zeta
potential and differential scanning calorimetry of native starch powders from
Dioscorea spp. 2 (2014) 632–638, https://doi.org/10.13171/mjc.2.5.2014.01.
01.17.
[25] R. Wongsagonsup, S. Shobsngob, B. Oonkhanond, S. Varavinit, Zeta potential (Z) and
pasting properties of phosphorylated or crosslinked rice starches, Starch/Staerke 57
(2005) 32–37, https://doi.org/10.1002/star.200400311.
[26] F. Xie, W.C. Liu, P. Liu, J. Wang, P.J. Halley, L. Yu, Starch thermal transitions compar-
atively studied by DSC and MTDSC, Starch/Staerke 62 (2010) 350–357, https://doi.
org/10.1002/star.200900258.
[27] B.R. Bhandari, T. Howes, Implication of glass transition for the drying and stability of
dried foods, J. Food Eng. 40 (1999) 71–79, https://doi.org/10.1016/S0260-8774(99)
00039-4.
[28] P. Liu, L. Yu, X. Wang, D. Li, L. Chen, X. Li, Glass transition temperature of starches
with different amylose/amylopectin ratios, J. Cereal Sci. 51 (2010) 388–391,
https://doi.org/10.1016/j.jcs.2010.02.007.
[29] S. Palzer, The effect of glass transition on the desired and undesired agglomeration
of amorphous food powders, Chem. Eng. Sci. 60 (2005) 3959–3968, https://doi.
org/10.1016/j.ces.2005.02.015.
[30] C. Turchiuli, A. Gianfrancesco, S. Palzer, E. Dumoulin, Evolution of particle properties
during spray drying in relation with stickiness and agglomeration control, Powder
Technol. 208 (2011) 433–440, https://doi.org/10.1016/j.powtec.2010.08.040.
[31] C.G. Biliaderis, Structural Transitions and Related Physical Properties of Starch,
Elsevier Inc., third ed., 2009 https://doi.org/10.1016/B978-0-12-746275-2.00008-2.
[32] M.T. Kalichevsky, E.M. Jaroszkiewicz, S. Ablett, J.M.V. Blanshard, P.J. Lillford, The glass
transition of amylopectin measured by DSC, DMTA and NMR, Carbohydr. Polym. 18
(1992) 77–88, https://doi.org/10.1016/0144-8617(92)90129-E.
[33] Y.P. Chang, P.B. Cheah, C.C. Seow, Plasticizing—antiplasticizing effects of water on
physical properties of tapioca starch films in the glassy state, J. Food Sci. 65
(2000) 445–451, https://doi.org/10.1111/j.1365-2621.2000.tb16025.x.
[34] R.M. Syamaladevi, S.K. Insan, S. Dhawan, P. Andrews, S.S. Sablani, Physicochemical
properties of encapsulated red raspberry (Rubus idaeus) powder: influence of
high-pressure homogenization, Dry. Technol. 30 (2012) 484–493, https://doi.org/
10.1080/07373937.2011.647369.
[35] J.A. Grabowski, V.D. Truong, C.R. Daubert, Spray-drying of amylase hydrolyzed
sweetpotato puree and physicochemical properties of powder, J. Food Sci. 71
(2006)https://doi.org/10.1111/j.1750-3841.2006.00036.x.
[36] J.D. Hoyos-Leyva, L.A. Bello-Pérez, J. Alvarez-Ramirez, E. Agama-Acevedo, Structural
characterization of aroid starches by means of chromatographic techniques, Food
Hydrocoll. 69 (2017) 97–102, https://doi.org/10.1016/j.foodhyd.2017.01.034.
[37] S.H. Hsu, T.R. Tsai, W.H. Chuo, T.M. Cham, Evaluation of Era-Tab as a direct compres-
sion excipient, Drug Dev. Ind. Pharm. 23 (1997) 711–716, https://doi.org/10.3109/
03639049709150773.
244 J. Hoyos-Leyva et al. / International Journal of Biological Macromolecules 120 (2018) 237–244
[38] H. Puchongkavarin, W. Bergthaller, S. Shobsngob, S. Varavinit, Characterization and
utilization of acid-modified rice starches for use in pharmaceutical tablet compres-
sion, Starch/Staerke 55 (2003) 464–475, https://doi.org/10.1002/star.200300232.
[39] V. Vongsurakrai, S. Varavinit, Product of spray dried rice starch (Era-Tab®) and its
utilization in pharmaceutical industry, Adv. Mater. Res. 93–94 (2010) 672–674,
https://doi.org/10.4028/www.scientific.net/AMR.93-94.675.
[40] M. Wootton, A. Bamunuarachchi, Water binding capacity of commercial produced
native and modified starches, Starch-Starke 30 (1978) 306–309, https://doi.org/
10.1002/star.19780300905.
[41] M. Lei, F.C. Jiang, J. Cai, S. Hu, R. Zhou, G. Liu, Y.H. Wang, H.B. Wang, J.R. He, X.G.
Xiong, Facile microencapsulation of olive oil in porous starch granules: fabrication,
characterization, and oxidative stability, Int. J. Biol. Macromol. 111 (2018)
755–761, https://doi.org/10.1016/j.ijbiomac.2018.01.051.
[42] X. Zeng, E. Ruckenstein, Supported chitosan-dye affinity membranes and their pro-
tein adsorption, J. Membr. Sci. 117 (1996) 271–278, https://doi.org/10.1016/0376-
7388(96)00079-8.