Available online at www.sciencedirect.

com

Journal of Food Engineering 88 (2008) 159–168

www.elsevier.com/locate/jfoodeng

Influence of glycerol and chitosan on tapioca starch-based

edible film properties
S. Chillo a, S. Flores b,c, M. Mastromatteo d, A. Conte e, Lı́a Gerschenson b,c,
M.A. Del Nobile a,e,*
a
Istituto per la Ricerca e le Applicazioni Biotecnologiche per la Sicurezza e la Valorizzazione dei Prodotti Tipici e di Qualità – BIOAGROMED,
University of Foggia, Via Napoli, 25, 71100 Foggia, Italy
b
Departamento de Industrias, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Argentina
c
Consejo National de Investigaciones Cientificas y Tecnicas (CONICET) de la Republica Argentina, Argentina
d
Department of Production Science, Engineering, Mechanics, Economics in Agriculture and Livestock Sistems, University of Foggia,
Via Napoli, 25, 71100 Foggia, Italy
e
Department of Food Science, University of Foggia, Via Napoli, 25, 71100 Foggia, Italy

Received 31 July 2007; received in revised form 18 January 2008; accepted 4 February 2008
Available online 9 February 2008

Abstract

The individual and interactive effects of glycerol and chitosan on tapioca starch-based edible film properties were investigated using
response surface methodology. Tests were run on the polymeric matrices to determine film forming solution apparent viscosity, mechan-
ical and dynamic-mechanical properties, water vapour permeability (WVP) and color. All film forming solutions exhibited pseudoplastic
behaviour. It was observed from the mechanical characteristics point of view, that the chitosan had a positive effect while the glycerol had
a negative effect. The tan d values were affected more by glycerol than the chitosan. With regards to WVP data, the chitosan addition had
a negative effect, whereas the glycerol one had a positive influence. Moreover, both the chitosan and glycerol influenced the color indices.
It can be concluded that the concentrations of chitosan and glycerol led to changes in tapioca starch edible film properties, potentially
affecting film performances.
Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Chitosan; Tapioca starch; Glycerol; Edible films; Response surface methodology

1. Introduction oil and moisture or perform as pouches or wraps. Among

Edible films and coatings have been particularly consid-
ered in food preservation, because of their capability in
improving global food quality (Franssen and Krochta,
2003; Franssen et al., 2002; Guilbert and Biquet, 1996;
Greener Donhowe and Fennema, 1994). The films can be
used to cover food surfaces, separate incompatible zones
and ingredients, form a barrier against oxygen, aroma,
*
Corresponding author. Address: Department of Food Science, Uni-
versity of Foggia, Via Napoli, 25, 71100 Foggia, Italy. Tel./fax: +39 881
589 242.
E-mail address: delnobil@unina.it (M.A. Del Nobile).
other important features, they can be used as carriers of
functional agents, as antioxidants or antimicrobials, and
to improve appearance and handling (Kester and Fennem-
a, 1986). Film production by natural and abundant biode-
gradable polymeric materials as cellulose, gums, starches or
proteins, is also convenient due to the lower environmental
consequences compared with common synthetic plastic
materials (Cutter, 2006).
Tapioca is a significant crop in South America, and it is
an economical source of starch (FAO, 2004). Tapioca
starch-based edible films exhibit appropriate physical char-
acteristics, since these film are odourless, tasteless, colorless
and impermeable to oxygen. However, films show

0260-8774/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfoodeng.2008.02.002
160 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168
brittleness with inadequate mechanical properties. The
physical properties of sorbate-containing tapioca starch–
glycerol edible films as affected by gelatinization and drying
process were studied by Flores et al., (2007a). They found
that the gelatinization technique and drying method used
to obtain edible films affected network characteristics deter-
mining changes in physical properties. Moreover, the
release of potassium sorbate from tapioca starch-based
edible films with glycerol as plasticizer was studied and
modelled by Flores et al. (2007b) who inferred that the pro-
posed model satisfactorily fits the experimental data.
Therefore, concluding that casting technique which pro-
duces films with high amorphous degree contributes greatly
to the matrix relaxation for sorbate release.
The addition of plasticizer agents is necessary to
improve the film flexibility. Glycerol is one of the most
popular plasticizers used in film-making techniques, due
to stability and compatibility with hydrophilic bio-poly-
meric packaging chain (Fernández Cervera et al., 2004).
The application of hydrophilic films, as starch-based films,
is limited by the water solubility and the poor water vapour
permeability. To solve this shortcoming, the blending of
starch with different bio-polymers (Xu et al., 2005) or the
addition of hydrophobic materials such as oils or waxes
(Garcı́a et al., 2000; Anker et al., 2001; Ayranci and Tunc,
2003) have been proposed.
The reasons for chitosan addition in edible films are the
good film forming and mechanical properties, no toxicity,
biodegradability, relative more hydrophobic nature that
could provide higher moisture barrier and water resistance
(Bangyekan et al., 2006; Mathew and Abraham, 2008). The
effects of acid (acetic, formic, lactic, propionic) concentra-
tions, plasticizers concentration and storage time on the
properties of solution-cast chitosan films have been widely
studied. Wong et al. (1992) investigated the water vapour
permeability of chitosan films using formic acid aqueous
solution. Caner et al. (1998) found that the tensile strength
was not time dependent, but elongation decrease with stor-
age time. The films formed with 7.5% lactic acid solution
had uniquely high values for the elongation at break and
the amount of added plasticizer affected the film properties.
Park et al. (2002) evaluated the characteristics of different
molecular weight chitosan observing that the mechanical
properties increased with chitosan molecular weight. More-
over, Hoagland and Parris (1996) prepared chitosan and
pectin laminated films with either the glycerol or lactic acid
as plasticizer. The films with chitosan and pectin had stor-
age and loss modulus significantly greater than the respec-
tive moduli of chitosan films alone. The water vapour
permeation of pectin or chitosan films, made with lactic
acid, was unchanged by lamination. The rheological and
surface tension properties of the filmogenic suspension
and water solubility, color, mechanical properties and
water vapour permeability of the composite film of chito-
san and methyl cellulose were determined by Garcı́a et al.
(2004). They found that the chitosan films showed rigid
characteristics (high elastic modulus and small elongation).
Bangyekan et al. (2006) evaluated film properties including
mechanical and physical properties of chitosan-coated cas-
sava starch film observing that an increase in chitosan coat-
ing led to a significant increase in tensile strength and
tensile modulus and a decrease in percentage elongation.
The study of properties of tapioca starch edible films in
relation to different concentrations of chitosan and glycerol
is a subject of great importance as it provides the knowl-
edge of the advantages and disadvantages of its possible
applications on the future of food shelf life.
The objective of this study was to determine the influ-
ence of chitosan and glycerol on the properties of tapioca
starch-based edible films, using an experimental design
and a response surface methodology analysis.
2. Material and methods
2.1. Preparation of edible films forming solutions
Chitosan solutions: chitosan of high molecular weight
(Sigma–Aldrich, St. Louis, MO) was dispersed in acetic
acid (J.T. Baker, Phillipsburg, NJ, USA) (1% v/v) to pre-
pare solutions of proper concentrations. The dispersions
were heated (80 °C) on a hotplate for 60 min under contin-
ous stirring to dissolve chitosan completely.
Tapioca starch solutions: Tapioca starch (Berasa S.A.,
Argentina) was dispersed in water–glycerol solutions in
order to obtain 4% (w/w) suspensions. The quantities of
glycerol were previously established. All dispersions were
heated (125 °C) on a hotplate for 30 min under continous
stirring to gelatinize starch completely.
Blend solutions: film-forming solutions were obtained by
mixing chitosan and starch/glycerol solutions in similar
proportions (50 g of each one) under gently stirring for
20 min. The chitosan and glycerol final contents of solu-
tions varied according to Table 1. Tapioca starch concen-
tration was maintained at 2% w/w in all blend solutions.
Finally, a vacuum was applied to remove air from the
systems.
2.2. Edible film preparation
Edible films were obtained by casting technique: 18 g of
film forming solutions were dispensed on the surface of
Petri dishes (9 cm diameter) and dried in a controlled tem-
perature chamber (32 °C) for 48 h. After film constitution,
Table 1
Variables and their levels for the CCD
Variable levels Independent variables
Chitosan (%) Glycerol (%)
2 0.1 0.5
1 0.325 0.688
0 0.55 0.875
+1 0.775 1.063
+2 1 1.25
 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168
it was separated from the dish to obtain stand-alone films.
All samples were conditioned inside desiccators containing
saturated saline solutions of NaBr which provided relative
humidity of 57.5–57.7% for 48 h at temperature of 25 °C,
before testing.
2.3. Rheological measurements
Rheological behaviour for each film forming solution
was investigated by means of a controlled-strain rotational
rheometer (ARES model, TA Instruments, New Castle,
DE, USA) equipped with a force rebalance transducer
(model 1K-FRTN1, 1–1000 g cm, 200 rad/s, 2–2000 gmf)
and a couette tool with concentric cylinder geometry (diam-
eter of cup and bob, 34 mm and 32 mm, respectively). A
steady temperature was ensured with an accuracy of
±0.1 °C by means of a controlled fluid bath unit and an
external thermostatic bath. Three repetitions for the flow
experiment were performed for each sample. Steady shear
stress over a range of shear rates of 50–1000 s1 was mea-
sured. Flow experiments were carried out at 25 °C. In order
to prevent water evaporation, a suitable cover tool sealing
the top of the couette tool was used during the test. The rhe-
ological behaviour was studied using the following power
law model that satisfactory fitted the experimental data:
s ¼ K  c_ n ; ð1Þ
where s is the shear stress (Pa), K is the consistency index
(Pa sn), the c_ is the shear rate and n is flow index (dimen-
sionless). The apparent viscosities were calculated at
300 s1 shear rate using the following power law model
(Bertuzzi et al., 2007):
g ¼ K  c_ n1 : ð2Þ
The fitting capability was calculated using the mean
square deviation (v2):
X ðxi  xi Þ2 ðf ðxi Þ  y Þ2
i
v2 ¼ 2
þ 2
; ð3Þ
i
r xi r yi
where (xi, yi) and ðxi ; f ðxi ÞÞ represent the measured data
point and the ‘nearest’ point on the function curve, respec-
tively. The mean square deviation is used when the mea-
surement errors are described by a Gaussian probability
distribution, and in this case the errors rxi and ryi corre-
spond to the standard deviation of the Gaussian
distributions.
2.4. Mechanical and dynamic-mechanical analyses
Both the mechanical and viscoelastic properties of the
films were determined using a stress-controlled Dynamic-
Mechanical Analyzer (DMA-Q 800, TA Instruments,
New Castle, DE, USA) equipped with a tension clamp.
Mechanical tests under static, transient and dynamic con-
ditions were performed, i.e. uniaxial tension, stress relaxa-
tion and oscillatory stress, respectively.
161
The dimensions of film samples used for the mechanical
and dynamic-mechanical tests were 20  5.6  0.08 mm
They were previously conditioned in desiccators containing
saturated saline solutions of NaBr that provided relative
humidity of 57.5–57.7% for 48 h at temperature of 25 °C.
Five repetitions were carried out for each test at room
temperature.
2.5. Unixial tension tests
With the aim to determine the elastic modulus (Ec), the
tensile strength and deformation at break of samples, the
stress–strain curves were acquired and analyzed. Tests were
carried out according to the following conditions: preload
force of 103 N, force ramp rate of 1 N min1.
The elastic modulus was evaluated from the initial slope
of the obtained stress–strain curve using the following
exponential equation (Del Nobile et al., 2007a,b):
rT ¼ Ec  eT  expðeT  KÞ; ð4Þ
where eN and rN are the true strain and the true stress,
respectively, calculated according to Mancini et al.
(1999); Ec is the elastic modulus (i.e. the tangent to the
stress strain curve at the origin); K is a constant value, re-
garded as a fitting parameter. Moreover, the tensile
strength and elongation at break were expressed as MPa
and percentage, respectively (Ninnemann, 1968).
2.6. Dynamic low-amplitude oscillatory tests
In order to determine the viscoelastic properties of
tested edible films, frequency sweep tests in tension were
performed, applying an oscillatory stress by using DMA-
Q 800; some fundamental dynamic-mechanical properties,
such as storage modulus (E0 ), loss modulus (E00 ) were deter-
mined in order to calculate complex modulus (E*). The
tan d value was calculated as E00 to E0 ratio (Steffe, 1996).
Instrumental settings were as follows: preload force
102 N and frequency range 0.02–200 Hz. The strain value
(0.1%) was obtained from preliminary strain sweep oscilla-
tory trials to determine the linear viscoelastic region. In
order to compare the E0 , E00 and tan d values, among the
investigated films, an oscillatory frequency of 50 Hz was
chosen as a reference.
2.7. Stress relaxation tests
Mechanical transient tests were performed to evaluate
the spectrum of the relaxation times from relaxation
curves. The relaxation data were obtained using DMA-Q
800 by setting the following experimental conditions:
preload force 102 N, quasi-instantaneous strain (1%), dis-
places time of 5 min. The relaxation modulus was calcu-
lated as the time-dependent stress divided by the imposed
strain. The relaxation behaviour of the investigated films
was described using Maxwell’s Generalized Model, accord-
ing to Bruckner et al. (2001):
162 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168
Z 1  t
rðtÞ
EðtÞ ¼ ¼ E0 þ EðkÞ  exp   dk
e0 0 k
where E(t) is the relaxation modulus expressed in MPa, i.e.
the elastic modulus at the decay time t; r(t) is the stress at
time t, expressed in MPa; e0 is the imposed strain; E(k)
(MPa) is the continuos distribution function (spectrum) of
relaxation time (the number of Maxwell elements ap-
proaches to infinity); k is the relaxation time in s; E0 is the
modulus of the single spring in parallel to Maxwell elements.
The relaxation spectrum is a fundamental quality in the lin-
ear theory of viscoelastic materials (Honerkamp and Weese,
1993) that does not depend on the experimental conditions
but only on the physical nature of the specimen. To describe
the continuous distribution function of relaxation time a
modified version of the expression proposed by Del Nobile
et al., (2007a,b), to represent the dynamic-mechanical behav-
iour of several viscoelastic foods, was used:
( The influence of chitosan and glycerol on tapioca starch-
 )
1 1 k 2
EðkÞ ¼ G1  0:5
exp  ð Þ : ð6Þ
d1  ð2  pÞ 2 d1
Eq. (6) is the normal distribution function with the mean
value equal to zero and the standard deviation equal to
d1, multiplied by a constant value (G1). The parameters
appearing in Eq. (6) account for the height (G1) and width
(d1) of the relaxation time distribution curve. By substitut-
ing Eq. (6) in Eq. (5) the following expression is obtained:
Z 1 ( "  2 #)
1 1 k
EðtÞ ¼ E0 þ G1  exp 
0 d1  ð2  pÞ0:5 2 d1
 t
 exp  dk ð7Þ
k i¼1
2.8. Color evaluation
Films disks of appropriate diameter were rested on a stan-
dard white background (Trezza and Krochta, 2000). Mea-
surements were performed by colorimeter (CR-400,
Minolta, Tokyo, Japan). The exposed area was relatively
greater than the illuminated area to avoid any light trapping
effect. The Hunter parameters: L, a and b, and the yellow
index (YI) were measured according to a standard test
method (ASTM D1925, 1995), in at least three randomly
selected positions for each sample. Color parameters range
from L = 0 (black) to L = 100 (white), a (greenness) to
+a (redness), and b (blueness) to +b (yellowness). Standard
values considered were those of the white background. Mea-
surements were made for C illuminant and 2° observer.
2.9. Water vapour permeability
Water vapour permeability (WVP) of films was deter-
mined gravimetrically at 25 °C using a modified ASTM
E96-00 (2000) procedure. The permeation cell (acrylic
cups) had an internal diameter (ID) of 4.4 cm and an exter-
nal diameter (ED) of 8.4 cm (exposed area: 15.205 cm2).
They had depth of 3.5 cm and contained CaCl2 (0% RH).
ð5Þ
Film was located between the cell and its acrylic ring
shaped cover that was adjusted to the cup with four screws.
A 10 mm air gap was left between the film and the CaCl2
layer. A rubber O-ring and vacuum grease helped to assure
a good seal. The covered cell was placed in a temperature
and RH controlled chamber (Ibertest, Daganzo, Madrid,
Spain) maintaining a temperature of 25 °C and a RH of
72–75%. After 20–24 h a stationary water vapour transmis-
sion rate was attained and, from that moment on, changes
in weight of the cell (to the nearest 0.1 mg) were recorded
daily over a 6 day period. All tests were conducted in trip-
licate and WVP values were calculated using the WVP Cor-
rection Method described by Gennadios et al. (1994).
2.10. Experimental design and statistical analysis
based film properties was studied by modulating the two
selected variables according to a two factor, five level
CCD (central composite design), as reported in Table 1.
Table 2 lists all experimental runs. The lowest and highest
levels of independent variables studied (chitosan and glyc-
erol percentage) and tapioca starch concentration were
chosen from results of preliminary laboratory tests.
A statistical software (Statistica for Windows V 7.1,
StatSoft, Tulsa, OK, USA) was used to fit a second-degree
polynomial function (Eq. (8)) to the independent variables
using the following equation:
X
2 X
2 2 X
X 2
w ¼ B0 þ ðBi  xi Þ þ ðBii  x2i Þ þ ðBij  xi  xj Þ;
i¼1 i¼1 j¼1
j6¼1
ð8Þ
where w is the generic dependent variable examined in this
work; B0 is the value of fitted response at the center point
of the design, i.e. (0, 0); Bi, Bj and Bij are the linear, qua-
dratic and cross-product regression terms, respectively; xi
and xj are independent variables (chitosan and glycerol)
that should affect w. Eq. (8) permitted evaluation of the ef-
fects of linear, quadratic and interactive terms of the inde-
pendent variables on the selected dependent variables. The
best-fit equations of the individual, quadratic and interac-
tive effects of chitosan and glycerol on some properties of
the edible film are reported in Table 3. Three-dimensional
plots (Iso-Response Surfaces) were generated by fixing
investigated variables at the center value of CCD. More-
over, the interactions among the apparent viscosity of the
film forming solutions and the mechanical and dynami-
cal–mechanical properties of developed films were evalu-
ated by multiple linear stepwise regression (p < 0.05).
The optimal composition of tapioca starch-based edible
film was selected on the basic idea reported by Wang et al.
(2006) to generate the optimum combination for whey pro-
tein isolate films. In this work, the desirability was not
determined instead the absolute residual error between
 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168 163

the observed and the predicted values of the dependent

3.258 ± 0.027
Composition of the 11 runs of the Central Composite Design, E0 , E00 values, fitting parameters (G1 and d1) from the monomodal Generalized Maxwell model, WVP values and color parameters of the

3.02 ± 0.27
3.59 ± 0.42
3.62 ± 0.30
5.48 ± 0.69
5.24 ± 0.27

4.90 ± 0.46
2.94 ± 0.33
7.10 ± 0.51
4.81 ± 0.35
5.10 ± 0.34
variables was calculated. The chosen optimal concentration
of glycerol and chitosan corresponded to the lowest abso-
lute residual error.

b
0.080 ± 0.047

0.097 ± 0.083

0.432 ± 0.049
0.122 ± 0.069
0.425 ± 0.071
0.058 ± 0.073
0.920 ± 0.089
0.502 ± 0.058
0.567 ± 0.071
0.033 ± 0.12

0.395 ± 0.18
3. Results and discussion

The influence of chitosan and glycerol concentrations

(denoted by XC and XG, respectively) on the properties
a

of tapioca starch-based edible film was analyzed in terms

Hunter scale

94.49 ± 0.29
93.94 ± 0.37
94.63 ± 0.59
92.91 ± 0.38
93.64 ± 0.50
95.38 ± 0.78
94.56 ± 0.49
95.12 ± 0.17
93.11 ± 0.39
94.58 ± 0.43
94.57 ± 0.31
of apparent viscosity of the film forming solutions,
mechanical and dynamical–mechanical characteristics of
the films, polymeric matrix color, water vapour permeabil-
L

ity and water solubility, as reported below. In the follow-

ing, each of the above aspect will be presented separately.
5.89 ± 0.56
7.02 ± 0.70
6.98 ± 0.62
10.45 ± 0.12
9.90 ± 0.53
6.23 ± 0.14
9.16 ± 0.87
5.69 ± 0.61
13.14 ± 0.99
8.92 ± 0.69
9.42 ± 0.63
3.1. Mechanical properties
YI

Fig. 1 shows an example of stress–strain curve relative to

1.73E10 ± 3.36E12
3.14E10 ± 2.07E11

2.75E10 ± 2.45E11
2.81E10 ± 5.95E11
1.58E10 ± 2.67E11
3.81E10 ± 3.31E11
3.5E10 ± 8.48E12
2.33E10 ± 4.23E11
3.1E10 ± 5.71E11
3.15E10 ± 2.93E11
1.49E10 ± 2.8E11

a combination (run 3) of the experimental design, where Ec

(MPa) is the slope of the initial curve.
WVP (g/Pa m s)

The best-fit equations describing the effects of the chito-

san and glycerol on both Ec values and tensile strength val-
ues were reported in Table 3; in the same table, the
summary of the multiple regression analysis was also
reported. As can be inferred from the above data, Ec values
26.28 ± 10.60

42.69 ± 17.96
44.71 ± 20.54

50.65 ± 20.76
67.36 ± 28.74
29.98 ± 10.19
61.81 ± 37.21

were positively affected by the linear and quadratic terms of

21.46 ± 1.40
37.03 ± 0.47

17.84 ± 1.18

17.48 ± 9.45

chitosan and glycerol contents, respectively, while nega-

tively influenced by glycerol and chitosan–glycerol interac-
tion. Also the tensile strength values were influenced by the
d1

individual positive term of the chitosan and by the negative

268.58 ± 104.03
92.09± 51.86

104.77 ± 38.92

187.99 ± 85.69

35.26 ± 23.54

influence of the interaction chitosan–glycerol. In order to

1.51 ± 0.51

7.09 ± 3.62
6.68 ± 3.08

0.67 ± 0.41
2.53 ± 1.50

5.51 ± 1.38

understand better the individual and interactive effects of

chitosan and glycerol variables on the mechanical parame-
ters, the surface response plots were analyzed. Fig. 2 shows
G1

the three-dimensional surface response plot of Ec vs the

97.52 ± 10.66
143.28 ± 31.74
149.43 ± 16.74
165.18 ± 14.28
147.72 ± 71.03
92.47 ± 15.97
110.75 ± 12.66
192.66 ± 18.12
143.58 ± 14.80
147.84 ± 22.47

two independent variables studied in this work. From the

176.62 ± 8.21

above figure it is possible to observe that the highest Ec val-

E00 (MPa)

ues of the composite film are obtained at the highest con-

centrations of chitosan and at the lowest concentrations
of glycerol. In fact, the highest Ec value was observed for
896.15 ± 112.34

959.16 ± 203.15
458.11 ± 140.70
619.85 ± 137.46
1151.50 ± 264.03

916.07 ± 170.37

546.73 ± 169.29

0.5% of glycerol and 1% of chitosan concentration. As

169.80 ± 39.24

141.05 ± 49.08

423.44 ± 64.63
227.49± 9.87

reported in literature, the plasticizers are intended to

decrease the intermolecular forces along polymer chains,
E0 (MPa)

imparting increased film flexibility while decreasing the

barrier properties of films (Banker, 1966; Caner et al.,
1998).
Glycerol (%)

Fig. 3 shows the surface plot of the influence of the inde-

pendent variables on the tensile strength values. From the
0.688
1.063
0.688
1.063
0.875

0.875
0.875
0.875
0.875
1.25

above figure it can be seen that the tensile strength value

0.5

increases with increasing the chitosan concentration. The

Chitosan (%)

above results are in accordance with the literature data per-

formed on composite film from 2% chitosan suspensions
(Garcı́a et al., 2004). On the other hand, a negligible effect
0.325
0.325
0.775
0.775
film samples

0.55
0.55
0.55

0.55
0.55
0.1

of the glycerol amount on the same dependent variable was

1
Table 2

measured. Although a negative influence of the chitosan–

Run

10
11

glycerol interaction on the tensile strength values was

1
2
3
4
5
6
7
8
9
164 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168

Table 3
Best-fit equations of the individual, quadratic and interactive effects of chitosan and glycerol on some properties of the edible film
Best-fit equations Fa SE R2
Ec (MPa) 9496.5 * [chitosan]  3888 * [glycerol] + 4315.1 * [glycerol]2  10378.9 * [chitosan] * [glycerol] 9.43 330.24 0.91
Elongation at 122.124 * [glycerol] 105.38 35.38 0.91
break (%)
Tensile strength 32.4285 * [chitosan]  25.4529 * [chitosan] * [glycerol] 63.73 1.98 0.93
(MPa)
E0 (MPa) 3122.72 * [chitosan]  2415.57 * [chitosan] * [glycerol] 79.41 173.65 0.94
E00 (MPa) 380.851 * [glycerol]  287.607 * [glycerol]2 + 83.627 * [chitosan] * [glycerol] 256.45 17.31 0.99
E* 3382.79 * [chitosan] + 253.15 * [glycerol]  3005.26 * [chitosan] * [glycerol] 91.75 143.62 0.96
tan d 0.200122 * [chitosan] + 0.291694 * [chitosan]2 + 0.740357 * [glycerol]2  0.972278 * [chitosan] * [glycerol] 659.61 0.02 0.99
E0 (MPa) 1987.70 * [chitosan]  855.63 * [glycerol] + 934.99 * [glycerol]2  2160.21 * [chitosan] * [glycerol] 7.32 75.93 0.90
d1 75.328 * [glycerol] + 131.125 * [chitosan]2  156.540 * [chitosan] * [glycerol] 21.54 16.08 0.88
YI 11.09 * [glycerol]  6.52 * [chitosan]2 + 8.30 * [chitosan] * [glycerol] 263.11 1.02 0.99
L 164.89 * [chitosan] + 107.05 * [glycerol]  187.72 * [chitosan] * [glycerol] 1030.16 5.61 0.99
a 0.707 * [chitosan] * [glycerol] 39.37 0.20 0.80
b 5.68 * [glycerol]  3.38 * [glycerol]2 +4.61 * [chitosan] * [glycerol] 209.71 0.60 0.99
g (Pa s) 1.57 * [chitosan] + 1.16 * [glycerol]2  2.39 * [chitosan] * [glycerol] 224.40 0.08 0.98
WVP (g/m s Pa) 1.91e10 * [chitosan] + 5.01e10 * [glycerol] 214.81 5.62e11 0.97
a
F, Fisher test value; SE, standard error; R2, regression coefficient.

Fig. 1. Stress plotted against strain for run 3 of the CCD.

Fig. 3. Effect of the interaction XC–XG on the independent variable

tensile strength of the tapioca starch-based edible films where tensile
strength is measured as MPa.

observed, results suggest that glycerol is mainly responsible

Fig. 2. Effect of the interaction XC–XG on the independent variable Ec of
the tapioca starch-based edible films where Ec is measured as MPa.
for the above behaviour. In fact, the higher tensile strength
value was observed for 0.5% and 1% of glycerol and chito-
san, respectively.
The elongation at break was influenced only by the glyc-
erol concentration. The best-fit equation describing effects
of the chitosan and glycerol concentration on elongation
at break values are reported in Table 3, along with the mul-
tiple regression analysis. As can be seen from the table the
elongation at break increased with the increase of glycerol
concentration. Xu et al. (2005) observed a dependence of
the elongation at break by starch to chitosan ratio with a
 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168
maximum value corresponding to a ratio of 1.5:1. Proba-
bly, in this study, the higher ratio starch to chitosan was
responsible for a negligible effect of the chitosan on the
elongation at break.
3.2. Dynamic-mechanical properties
Table 2 reports the E0 and E00 values. As can be inferred
from the above data, the glycerol concentration had a neg-
ative effect on the E0 values and a positive influence on the
E00 values. On the other hand, the chitosan percentage
affects positively both E0 and E00 values.
The best-fit equation describing effects of the chitosan
and glycerol concentration on tan d and E* values are
reported in Table 3, along with the multiple regression
analysis. The tan d values are positively affected by the indi-
vidual term of chitosan, the quadratic term of chitosan and
glycerol, and negatively affected by the interaction chito-
san–glycerol. For the sake of clarity Fig. 4 shows the
three-dimensional surface response plot of tan d vs the
investigated independent variables. As can be inferred from
the above figure, tan d values increase with increasing glyc-
erol concentration, whereas a slight decrease is observed as
the chitosan concentration decreases. Moreover, it can be
observed that the minimum tan d value is that correspond-
ing to low concentrations of glycerol and chitosan. The
minimum tan d value denotes a prevalently elastic behav-
iour with E0 higher than the E00 .
E* was affected by the individual positive terms of chito-
san and glycerol and by the negative interaction chitosan–
glycerol. Therefore, the E* values increased with the
increase of the chitosan and with the decrease of the glyc-
Fig. 4. Effect of the interaction XC–XG on the on the independent
variable tan d of the tapioca starch-based edible films.
165
erol. An increase of the glycerol concentration determined
a slight increase of the E* values.
3.3. Stress relaxation
As reported above the viscoelastic properties of poly-
meric materials can be expressed by different fitting param-
eters (G1, d1 and E0). Fig. 5 shows stress-relaxation results
for a test run (6) which fitted to Eq.(7). The curve shown in
the above figure was obtained by fitting Eq. (7) to the
experimental data. As can be seen from the figure the
monomodal Generalized Maxwell model excellently fits
the experimental data (v2 = 0.037). The calculated values
of G1 and d1 from the monomodal Generalized Maxwell
model were listed in Table 2. The data highlight that an
increase in the chitosan concentration and a decrease in
the glycerol concentration bring about an increase in G1
values. On the other hand, by increasing glycerol concen-
tration higher d1 values were obtained and by increasing
chitosan concentration lower d1 values were obtained.
The best-fit equations describing effects of chitosan and
glycerol concentrations on the constant E0 values are
reported in Table 3, along with the multiple regression
analysis. The values of this parameter were positively
affected by the individual term of the chitosan and the glyc-
erol quadratic term and negatively by the individual term
of the glycerol and the interaction chitosan–glycerol.
Fig. 6 shows the three-dimensional surface response plot
of E0 vs chitosan and glycerol. From the above figure, it
can be observed that the E0 values increased with increas-
ing chitosan and decreasing glycerol concentration.
3.4. Apparent viscosity
Fig. 7 was obtained by fitting Eq. (1) to the experimental
data (v2 = 0.311). The parameters K (consistency index)
and n (flow index) were determined in order to calculate
the apparent viscosity to a selected value of shear rate
(300 s1) by using Eq. (2). The film forming solutions
showed a non-Newtonian behaviour. Moreover, their
Fig. 5. E(t) plotted against time for run 6 of the CCD; (s) experimental
data, best fit (—).
166 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168

Table 4
Multiple linear stepwise regression to evaluate the interactions among
apparent viscosities of the film forming solutions and mechanical and
dynamical–mechanical film properties
Variables Coefficient of Standard t p-Level
regression (B) error (SE)
Ec 0.43178 0.071617 –6.02904 0.001807
Tensile strength 17.95176 3.327484 5.39499 0.002954
Elongation at break 0.59431 0.240001 –2.47629 0.056096
E0 0.24406 0.104235 –2.34142 0.066263
E00 1.34033 0.480930 2.78695 0.038584
G1 – – – –
d1 – – – –
E0 1.93612 0.327570 5.91056 0.001974
–: Variables with null contribution.
p < 0.00013, R2 = 0.99, Adj. R2 = 0.97, SE = 13.03, F = 66.519,
Accuracy = 0.07.

accuracy coefficient of the model was also reported (Ross,

1996). In the table it can be observed that the apparent vis-
cosities of the film forming solutions show a positive
Fig. 6. Effect of the interaction XC–XG on the independent variable E0 of regression coefficient for the tensile strength, E00 and K,
the tapioca starch-based edible films where E0 is measured as MPa. whereas Ec showed a negative regression coefficient.

3.5. Water vapour permeability

The best equation that describes the relationship

Fig. 7. Shear stress plotted against shear rate for run 6 of the CCD; (s)
experimental data, best fit (—).
apparent viscosity was highly dependent on the shear rate
(Steffe, 1992). In particular, all film forming solutions
presented pseudoplastic behaviour (n < 1). The best-fit
equations describing effects of the chitosan and glycerol
concentrations on the apparent viscosity g are reported in
Table 3, along with the multiple regression analysis. The
values of this parameter were positively affected by the
individual term of the chitosan and by the positive
quadratic effect of the glycerol. Moreover, a negative inter-
action chitosan–glycerol was observed. Therefore, the
apparent viscosity increased with the increase of the glyc-
erol and with the decrease of the chitosan.
Table 4 reports the results of the multiple linear stepwise
regression to evaluate the interactions among apparent vis-
cosities of the film forming solutions and films mechanical
and dynamical–mechanical properties. In this table the
between water vapour permeability and the independent
factors (chitosan and glycerol) is shown in Table 3. Table
2 reports the water vapour permeability of the film sam-
ples. It can be seen that chitosan addition had a negative
effect on water vapour permeability values of the film,
whereas glycerol had a positive influence. In both cases,
only the linear contribution is statistically significant. These
tendencies are according to previous results of other
authors. Arvanitoyannis et al. (1998) attributed the perme-
ability increment to plasticizing effect of glycerol, which
reduce polymeric packaging density. On the other hand,
the increase of interactions between chitosan and tapioca
starch (hydrogen bonding type) with chitosan concentra-
tion, reduces the hydrophilic group availability reducing,
therefore, the water vapour transmission rate (Xu et al.,
2005). The WVP results reveal that permeability is strongly
influenced by glycerol, being higher as plasticizer concen-
tration increase. On the other hand, permeability is slightly
and negatively affected by chitosan.
3.6. Color evaluation
Color attributes are of prime importance because they
directly influence consumer acceptability. Chitosan–starch
films, containing different amounts of glycerol, resulted
homogeneous and transparent materials. The color attri-
butes obtained for each film sample are summarized in
Table 2. The results show that YI, b and L parameter of
Hunter scale increased with chitosan concentration. More-
over, a values of Hunter scale increased with the increase of
the glycerol and the decrease of the chitosan. These results
 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168 167

Table 5 Acknowledgments
Optimal composition of tapioca starch-based edible film
Dependent Goal Observed Predicted Absolute residual We acknowledge the financial support from Universidad
variables value value errora (%) de Buenos Aires, Consejo Nacional de Investigaciones
Ec (MPa) Maximum 2031.93 1503.65 25.99 Cientı́ficas y Técnicas de la República Argentina, Agencia
Tensile strength Maximum 10.87 10.84 0.28 Nacional de Investigaciones Cientı́ficas y Tecnológicas de
(MPa)
Elongation at Maximum 166.65 129.82 22.10
la República Argentina and Secretarı́a de Ciencia, Tec-
break (%) nologı́a y Innovación Productiva and Ministerio de Asun-
tan d Minimum 0.13 0.12 7.69 tos Externos of Italy (SECyT-MAE).
WVP (g/m s Pa) Minimum 1.49E10 1.77E10 18.79
a
The absolute residual error has been computed as: [(Observed value 
Predicted value)/Observed value] * 100. References

suggest that the presence of investigated bio-polymer gives
origin to a darker film. It was also observed that, in
general, the glycerol increase produced an increase in YI,
b values and L. Similar results were reported by Garcı́a
et al., 2004.
3.7. Optimal composition of tapioca starch-based edible film
The results of the statistical analysis leading to the opti-
mal composition of tapioca starch-based edible film were
reported in Table 5. The absolute residual error for the
dependent variables Ec, tensile strength, elongation at
break, tan d and WVP ranged from 0.28% to 25.99%. The
tapioca starch-based edible film optimization could be
achieved using the composition corresponding to the low-
est absolute residual errors (0.28%, 7.69%) of the dependent
variables. This composition contains values of chitosan
and glycerol concentrations equal to 0.55% and 0.5%,
respectively.
4. Conclusions
The properties of starch-based edible films were influ-
enced by chitosan and glycerol concentrations. In partic-
ular, Ec and tensile strength were positively influenced by
chitosan and negatively by chitosan–glycerol interaction.
For WVP, the chitosan presence had a negative effect,
whereas the glycerol concentration had a positive influ-
ence. The glycerol concentration had no effect on the
E0 values and a positive influence on the E00 values;
whereas, the chitosan concentration affected positively
both E0 and E00 values. Moreover, tan d values increased
with increasing glycerol concentration, and a slight
decrease was observed as the chitosan concentration
decreased. The increasing chitosan addition produced
an increase of YI, b parameter of Hunter scale and L
values; moreover, a values of Hunter scale increased with
the decrease of the chitosan. Instead, increasing glycerol
concentration determined an increase in YI, b, a and L
values. The optimal composition of tapioca starch-based
edible films investigated using response surface methodol-
ogy corresponded to chitosan and glycerol concentrations
equal to 0.55% and 0.5%, respectively.
Anker, M., Berntsen, J., Hermansson, A.M., Stading, M., 2001. Improved
water vapour barrier of whey protein films by addition of an acetylated
monoglyceride. Innovative Food Science and Emerging Technologies
3, 81–92.
Arvanitoyannis, S.I., Nakayama, A., Aiba, S., 1998. Chitosan and gelatin
based edible films: state diagrams, mechanical and permeation
properties. Carbohydrate Polymers 37, 371–382.
ASTM D1925, 1995. Standard Test Method for Yellowness Index of
Plastics. American Society for Testing and Materials, Philadelphia.
ASTM E96-00, 2000. Standard Test Method for Water Vapour Trans-
mission of Materials. American Society for Testing and Materials,
Philadelphia.
Ayranci, E., Tunc, S., 2003. A method for the measurement of the oxygen
permeability and the development of edible films to reduce the rate of
oxidative reactions in fresh foods. Food Chemistry 80, 423–431.
Bangyekan, C., Aht-Ong, D., Srikulkit, K., 2006. Preparation and
properties evaluation of chitosan-coated cassava starch films. Carbo-
hydrate Polymers 63, 61–71.
Banker, G.S., 1966. Film coating theory and practice. Journal of
Pharmaceutical Science 55, 81–89.
Bertuzzi, M.A., Armada, M., Gottifredi, J.C., 2007. Physicochemical
characterization of starch based films. Journal of Food Engineering 82,
17–25.
Bruckner, S., Allegra, G., Pegoraro, F., La Mantia, F., 2001 (Proprietà
viscoelastiche dei polimeri). Scienza e Tecnologia dei Materiali
Polimerici, 1nd ed. Edises Press, Napoli, Italia.
Caner, C., Vergano, P.J., Wiles, J.L., 1998. Chitosan film mechanical and
permeation properties as affected by acid, plasticizer, and storage.
Journal of Food Science 63 (6), 1049–1053.
Cutter, C.N., 2006. Opportunities for bio-based packaging technologies to
improve the quality and safety of fresh and further processed muscle
foods. Meat Science 74, 131–142.
Del Nobile, M.A., Chillo, S., Falcone, P.M., Laverse, J., Pati, S., Baiano,
A., 2007a. Textural changes of Canestrello Pugliese cheese measured
during storage. Journal of Food Engineering 83, 621–628.
Del Nobile, M.A., Chillo, S., Mentana, A., Baiano, A., 2007b. Use of
the generalized Maxwell model for describing the stress relaxation
behavior of solid-like foods. Journal of Food Engineering 78, 978–
983.
FAO, Food and Agriculture Organization, 2004. Proceedings of the
Validation Forum on the Global Cassava Development Strategy. In:
Global Cassava Market Study Business Opportunities for the Use of
Cassava, vol. 6. International Fund for Agricultural Development,
Roma, Italia.
Fernández Cervera, M., Karjalainen, M., Airaksinen, S., Rantanen, J.,
Krogars, K., Heinämäki, J., Iraizoz Colarte, A., Yliruusi, J., 2004.
Physical stability and moisture sorption of aqueous chitosan–amylose
starch films plasticized with polyols. European Journal of Pharma-
ceutics and Biopharmaceutics 58, 69–76.
Flores, S., Famá, L., Rojas, A.M., Goyanes, S., Gerschenson, L., 2007a.
Physical properties of tapioca-starch edible films: influence of film-
making and potassium sorbate. Food Research International 40, 257–
265.
168 S. Chillo et al. / Journal of Food Engineering 88 (2008) 159–168
Flores, S., Conte, A., Campos, C., Gerschenson, L., Del Nobile, M.A.,
2007b. Mass transport properties of tapioca-based active edible films.
Journal of Food Engineering 81, 580–586.
Franssen, L.R., Krochta, J.M., 2003 (Edible coatings containing natural
antimicrobials for processed foods). In: Soller, S. (Ed.), Naturals
Antimicrobials for the Minimal Processing of Foods. CRC Press, Boca
Ratón, FL, USA.
Franssen, L.R., Rumsey, T.R., Krochta, J.M., 2002. Modeling of
natamycin and potassium sorbate diffusion in whey protein
isolate films for application to cheddar cheese. Poster 28-5,
Institute of Food Technologists Annual Meeting, Anaheim,
California.
Garcı́a, M., Martino, M., Zaritizky, N., 2000. Lipid addition to improve
barrier properties of edible starch-based films and coatings. Journal of
Food Science 65 (6), 941–947.
Garcı́a, M.A., Pinotti, A., Martino, N.M., Zaritzky, N.E., 2004. Char-
acterization of composite hydrocolloid film. Carbohydrate Polymers
56, 339–345.
Gennadios, A., Weller, C.L., Gooding, C.H., 1994. Measurement errors in
water vapour permeability of highly permeable, hydrophilic edible
films. Journal of Food Engineering 21, 395–409.
Greener Donhowe, I., Fennema, O., 1994 (Edible films and coatings:
characteristics, formation, definitions, and testing methods). In:
Krochta, J.M., Baldwin, E., Nı́speros-Carriedo, M.O. (Eds.), Edible
Films and Coatings to Improve Food Quality. Technomic Publishing
Co. Inc., Lancaster, PA.
Guilbert, S., Biquet, B., 1996 (Edible films and coatings). In: Bureau, G.,
Multon, J.L. (Eds.), In: Food Packaging Technology, vol. 1. Wiley-
VCH Inc., New York, USA, pp. 315–353.
Hoagland, P., Parris, N., 1996. Chitosan/pectin coated films. Journal of
Agricultural and Food Chemistry 44, 1915–1919.
Honerkamp, J., Wesse, J., 1993. A nonlinear regularization method for the
calculation of relaxation spectra. Rheological Acta 32, 65–73.
Kester, J.J., Fennema, O.R., 1986. Edible films and coatings: a review.
Food Technology 12, 47–59.
Mancini, M., Moresi, M., Rancini, R., 1999. Mechanical properties of
alginate gels: empirical characterization. Journal of Food Engineering
39, 369–378.
Mathew, S., Abraham, T.E., 2008. Characterisation of ferulic acid
incorporated starch–chitosan blend films. Food Hydrocolloids 22,
826–835.
Ninnemann, K.W., 1968 (Measurement of physical properties of flexible
films). In: Sweeting, O.J. (Ed.), The Science and Technology of
Polymer Films. Interscience, London, England, pp. 546–649.
Park, S.Y., Marsh, K.S., Rhim, J.W., 2002. Characteristics of different
molecular weight chitosan films affected by the type of organic
solvents. Journal of Food Science 67, 194–197.
Ross, T., 1996. Indices for performance evaluation of predictive models in
food microbiology. Journal of Applied Bacteriology 81 (5), 501–508.
Steffe, J.F., 1992. Rheological Methods in Food Process Engineering.
Freeman Press, Michigan, pp. 17–27.
Steffe, J.F., 1996. Rheological Methods in Food Process Engineering,
second ed. Freeman Press, Michigan.
Trezza, T.A., Krochta, J.M., 2000. Color stability of edible coatings
during prolonged storage. Journal of Food Science 65 (1), 1166–1169.
Wang, L., Liu, L., Holmes, J., Huang, J., Kerry, J.F., Kerry, J.P., 2006.
Effect of pH and addition of corn oil on the properties of whey protein
isolate-based films using response surface methodology. International
Journal of Food Science and Technology. Accepted in revised form 21
November 2006.
Wong, D.W.S., Gastineau, F.A., Gregorski, K.S., Tillin, S.J., Pavlath,
A.E., 1992. Chitosan-lipid films: microstructure and surface energy.
Journal of Agricultural and Food Chemistry 40, 540–544.
Xu, Y.X., Kim, K.M., Hanna, M.A., Nag, D., 2005. Chitosan–starch
composite film: preparation and characterization. Industrial Crops and
Products 21, 185–192.