Polymer Testing: Sciencedirect
Polymer Testing: Sciencedirect
Polymer Testing: Sciencedirect
Polymer Testing
journal homepage: www.elsevier.com/locate/polytest
Keywords: In the present study, new biodegradable nanocomposite films were produced. The effect of wheat gluten (WG)
Gluten (1–2.5 wt%), carboxymethyl cellulose (CMC) (0.5–1 wt%) and cellulose nanofiber (CNF) (0-10 wt% based on
Carboxymethyl cellulose gluten and CMC biopolymers solid matter in each formula) concentration on water vapor permeability (WVP),
Cellulose nanofiber mechanical and color properties of the biodegradable nanocomposites were investigated using response surface
Nanocomposites
methodology (RSM). Regression models were also developed for all responses as a function of linear, interaction
Response surface methodology
and quadratic terms of WG/CMC/CNF concentrations. The optimum point with 1 wt% gluten, 0.686 wt% CMC
and 8.549 wt% CNF for a maximum desirability of 0.878 was obtained. Furthermore, the microstructure, surface
topography and the distribution quality of cellulose nanofiber in the biopolymer matrix were evaluated by
scanning electron microscopy (FE-SEM), atomic force microscopy (AFM) and X ray diffraction (XRD) analysis.
1. Introduction at dry state [6]. The wheat gluten film is brittle and therefore, several
methods have been used for modifying it including addition of plasti-
Synthetic polymers are widely used as food packaging materials due cizers and cross-linking agents. However, similar to the most of other
to its low cost and easy availability. Because of the environmental biopolymers, the relatively poor mechanical properties (poor toughness
concerns associated with the wastes of synthetic packaging, consider- and relatively high brittleness) and water sensitivity limit its applica-
able efforts have been made to develop new biodegradable packaging tions as high performance packaging polymer [5–7].
[1]. In this regard, two types of biodegradable polymers have been Carboxymethyl cellulose (CMC) is one of the most commonly used
developed: edible and non-edible biopolymers [2]. In recent years, at- polysaccharide for making biodegradable films due to its high viscosity,
tention has been given to biodegradable polymers which can be ob- low cost, non-toxicity and good film properties (high clarity and me-
tained from by-products of food industries. Common bio-based mate- chanical strength) [9]. However, the numerous hydroxyl and carboxylic
rials which potentially can be used in bioplastic production include groups in CMC contribute to high water binding and moisture sorption
carbohydrates, proteins, lipids and combination of them [3]. properties [10]. It can be used with other biopolymer such as WG for
Wheat gluten (WG) is one of the most promising renewable re- producing of blend films and improving of physico-mechanical prop-
sources for the production of bioplastic due to having relatively low erties [11].
price, high abundant, fast biodegradability, and good film-forming Cellulose nanostructures can be used as high performance filler for
properties [4]. In addition, it is available as a byproduct from the starch bioplastic due to their suitable properties, e.g. biodegradability, having
industry [5]. Wheat gluten films exhibit effective barrier properties a natural, abundant source and reasonable price, low density, and high
against gases (such as oxygen, carbon dioxide and aroma compounds) mechanical strength [12,13]. Cellulose nanostructures have two forms
*
Corresponding author. Department of Food Science and Technology, Faculty of Agriculture, University of Tabriz, P.O. Box 51666-16471, Tabriz, Iran.
**
Corresponding author.
E-mail addresses: [email protected], [email protected] (B. Ghanbarzadeh), [email protected] (A. Ayaseh).
https://doi.org/10.1016/j.polymertesting.2019.105989
Received 18 May 2019; Received in revised form 5 July 2019; Accepted 13 July 2019
Available online 15 July 2019
0142-9418/ © 2019 Published by Elsevier Ltd.
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
including cellulose nanocrystals (CNCs) and cellulose nanofibers (CNFs) average diameter of 35 nm was obtained from Nano Novin polymer Co,
[14]. Several researches have been reported that incorporation of CNFs (Mazandaran, Iran). Glycerol and other chemical materials used in this
in biopolymer matrixes can potentially improve mechanical and barrier study with analytical grade were purchased from Merck co (Darmstadt,
properties of the resulted nanocomposite films [15]. Germany).
The great majority of biopolymer-based films lose their mechanical
and barrier properties upon hydration and the main functional prop-
erties of these hydrophilic polymers strongly depend on their water 2.2. Preparation of WG/CMC/CNF nanocomposite films
content and therefore on the surrounding humidity [16]. Addressing
these drawbacks is a key factor in the study of bio-based materials and The nanocomposite film was prepared by casting method. The
the development of novel bioplastics and so, a number of approaches gluten solution was prepared according to a formerly described method
have been considered for improving undesirable properties of them. by Kayserilioğlu et al. [18], with some minor modifications. First, the
The most important approaches include blending of different biopoly- distilled water pH was adjusted to 11–11.2 by adding sodium hydroxide
mers (blend films), mixing with lipids (emulsified films), adding of solution. Then, gluten (1–2.5 g) was dispersed in 50 ml adjusted dis-
cross linking agent and using of nano fillers in polymer matrix (nano- tilled water from the previous stage, using through a sieve, and was
composite films) [11]. Incorporation of the nanoparticles (such as stirred at low speed. The pH of the gluten suspension was adjusted and
montmorillonite, metal oxides, chitin and cellulose nano crystals) not kept at 11–11.2 using 0.5 M NaOH solution. After 30 min of stirring the
only potentially improves physical-mechanical features of biopolymer gluten solution heated for 10 min at 70 °C using a water bath. Then,
based films but also may add antibacterial and nutritional properties to CMC (0.5–1 g) was dissolved in a separate beaker with 35 mL of dis-
them [6]. tilled water using magnetic stirring and heated in a water bath at 75 °C
Response surface methodology (RSM) is an empirical modeling for 20 min. Also, cellulose nanofiber (0–10% based on solid matter) was
method which is used to assessment of several independent variables dispersed in 15 ml distilled water and was sonicated for 20 min in an
effect on the response with the goal of optimizing the response [17]. ultrasonic bath. After that, the different film forming solutions were
RSM modeling can provide more accurate and complete data by doing mixed together. Finally, glycerol as plasticizer (30% based on the bio-
minimal number of experimental so that it has many advantages in polymer) was added to the solution and mixed slowly for 20 min for
optimizing the film formulation. removing the air bubbles. Then, 50 ml of the film-forming solution was
According to our knowledge, the effect of CNF on the WG-CMC casted on the teflon mold (14 cm diameter) and dried in incubator at
based films has not been investigated. In the present research, we stu- 30 °C for 24 h. The dried films were peeled off from the mold, and
died the optimization of CNF, WG and CMC levels by the RSM method stored in a polyethylene bags at 4 °C (Fig. 1).
for obtaining to bionanocomposite film with suitable physico-chemical
properties.
2.3. Characterization of the films
2. Material and methods
2.3.1. Determination of Mechanical properties
2.1. Materials Mechanical properties of nanocomposite film ultimate tensile
strength (UTS), percentage of strain at break (SB) and Young's modulus
Wheat gluten with was obtained from Shahdineh Aran Co, (Esfahan, (YM) were determined following a standard method of ASTM (Standard
Iran) and the protein content of the wheat gluten were calculated ac- D638, 2010) on the conditioned films at 55 ± 3% RH and 25 °C for
cordance to ISO 1871 standard (79–81% protein). Carboxymethyl cel- 24 h using a Sanaf Universal Testing Machine (Tehran, Iran) with a 25 N
lulose (CMC), with a degree of substitution of (0.82) and viscosity of load cell and crosshead speed of 5 mm/min at room temperature. The
(2771 cps) was purchased from Shandong young cellulose technology specimens cut in to dumbbell shape with special cuter and then loaded
co, (Linyi, Shandong, China). Cellulose nanofiber in the form of gel with an initial grip separation of 50 mm. Each specimen tested at least 3
(2.5%), produced by mechanical synthesis, purity of 99% ≤ and replicated and their average reported.
2
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
2.3.2. Determination of water vapor permeability (WVP) recorded at scan rate of 1° min−1 and a step size of 0.02, and test was
The water vapor permeability (WVP) rate of the nanocomposite carried out at room temperature.
films was calculated according to ASTM E96-95 standard method. All
film samples were conditioned at 55 ± 3% RH at 25 °C for 24 h. After, 2.4. Experimental design and optimization
each test cup containing 3 gr anhydrous calcium sulfate (0% RH) were
covered with film samples. Then, all test cups were placed in a de- Film concentration ratios were optimized to investigate the sy-
siccator containing saturated potassium sulfate solution (97 ± 2% RH) nergistic effect of the variables using the central composite experi-
at 25 °C. The water vapor transmission rate of nanocomposite films was mental design (CCD). RSM in order to survey the effect of gluten con-
studied from the weight gain and absorbed cups over different times. centration (X1) (1–2.5 wt%), CMC concentration (X2) (0.5–1 wt%) and
After that, slope was obtained by linear regression from the weight and cellulose nanofiber (X3) (0-10 wt%) on the depended response variables
time changes. Finally, water vapor permeability (WVP) of nano- included water vapor permeability (WVP), ultimate tensile strength
composite films was calculated from the following equation. (4): (UTS), strain at break (SB), Young's modulus (YM), total color differ-
WVTR. X ence (ΔE), yellowness index (YI), and whiteness index (WI) of the edible
wvp = films were used. The suitable levels of the independent variables fol-
P (R2 R1) (4)
lowing preliminary tests to obtain the appropriate mechanical proper-
Here, the water vapor transmission rate (WVTR) was defined as the ties, barrier and color properties of the composite film also acceptability
slope (g/h) specified by the transfer area (m2). X is the average film cohesion were determinates. All responses and film preparation con-
thickness (mm). P is the saturated vapor pressure of water (Pa) at the ditions for each formulation type was independently repeated three
test temperature (25 °C), R2 is the RH in desiccator, R1, amount of RH in times.
the test cup. All measurements were carried out at least 3 replicates. Generally, 20 formulations with five replicates at the center point in
order to determined pure error and repeatability of all data according to
2.3.3. Determination of film color a central composite design were obtained. Table 1 shows the complete
The color properties of the biodegradable nanocomposite films was experimental design used for gluten, CMC and CNF nanocomposites and
analyzed with a colorimeter (made in faculty of agriculture, University the coded and uncoded levels of the independent variables. The ex-
of Tabriz, Iran). Film specimens were placed on a white standard plate perimental design arrangement was randomized to minimize the effects
(L = 96.86, a = −0.02, and b = 1.99) and the lightness (L), “a” (red- of uncontrolled factors. Data were fitted to a second order equation as a
green) and “b” (yellow blue) color values were measured. “L” values function of dependent variables Eq. (5):
range from 0 (black) to 100 (white); “a” values range from −80 2 2
(greenness) to 100 (redness); and “b” values range from −80 (blueness) Y= 0 + 1 X1 + 2 X2 + 3 X3 + 12 X1 X2 + 13 X1 X3 + 11 X1 + 22 X2
3
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
2.5. Verification and optimization procedures were 0.9842, 0.9864, 0.9855, 0.9677, 0.9564, 0.9673 and 0.9499
while, the R2adj values were 0.9699, 0.9741, 0.9725, 0.9386, 0.9172,
The numerical and graphical optimization procedures in order to 0.9378 and 0.8953, respectively (Table 3). No considerable difference
obtain the optimum level of three independent variables (X1: gluten, between R2 and R2 adj values indicated that most of model terms were
X2: CMC and X3: CNF) in order to specify the most appropriate value of significant and selected polynomial model had the adequate accuracy
responses and to achieve the best nanocomposite film properties in- [23]. The coefficient of variation (CV) values was less than 10% for
cluding WVP (g/mhpa), UTS (MPa), SAB (%), Young's modulus (YM) UTS, SB, ΔE, WI, YI and WVP. The adequate precision value, which
and color properties (ΔE, L and YI) were applied. The experimental data measures the signal to the noise ratio, was found to be greater than 4 for
and fitted values predicted by the models were compared in order to all responses which indicate an adequate signals and shows model can
determine the validity and adequacy of the regression models. The be used to navigate the design space. In order to develop the simple
numerical and graphical optimizations were also performed by Design- response surface models, all insignificant terms (P > 0.05) were
Expert 10.0.7 software. eliminated (Table 4). After excluding non-significant factors, the final
regression equations (in terms of uncoded factors 6-12) to predict the
3. Results and discussion effects of gluten, CMC and CNF concentrations on the responses were
shown in equations (6)-(12).
3.1. Optimization of gluten/CMC/CNF composite films WVP = 1.91 + 6.51X1 + 8.11X2 0.02X3 (6)
3.1.1. Response surface model analysis UTS = 41.07 + 10.41X1 + 1.02X3 11.83X1 X2 63.68X22 0.06X32
In this study, RSM was used to study linear, quadratic and inter- (7)
action effects of gluten, CMC and CNF concentrations (independent
variable) on each studied response (dependent variable). SB = 532.35 180.02X1 + 7.78X3 + 108.67X1 X2 8.56X2 X3 + 36.06X12
In order to evaluate the model's adequacy, different parameters + 463.02X22 (8)
including the model F value, the lack of fit F value, the coefficient of
determination R2, adjusted R2, press value and coefficient of variation YM = 1039.23 + 295.66X1 + 2454.06X2 297.22X1 X2 10.41X1 X3
(CV) were performed. The results of experimental data obtained by the 1374.65X22 2.04X32 (9)
response variables and films thickness are shown in Table 2.
The quadratic models were significant (P < 0.01) for all responses E = 55.09 17.27X1 82.17X2 + 13.42X1 X2 + 0.45X1 X3 + 4X12
(WI, YI, AE, UTS, YM, SB and WVP). The lack of fit values of quadratic
+ 46.96X22 + 0.04X32 (10)
model was non-significant (P < 0.01) for all responses.
In Table 3, the amount and p-value of linear (X1, X2, X3), quadratic
YI = 142.6 50.06X1 185.77X2 + 34.88X1 X2 + 1.27X1 X3 + 11.78X12
(X11, X22, X33), interactive (X1X2, X1X3, X2X3) coefficients, model p-
value and F-value has been presented. Lower p-value indicates higher + 100.76X22 + 0.15X32 (11)
significance of the corresponding coefficient and the p-value lower than
0.01 shows that the corresponding coefficient is highly significant and WI = 8.73 + 20.88X1 + 67.93X2 –9.71X1 X2 –0.24X1 X3 4.86X12 37.69X22
suitable for application in regression model. 0.07X32 (12)
The R2 and adj-R2 shows measure of the quality of fitting of ex-
Final regression equations (in terms of coded factors 13-19) were as
perimental data on the model and adequacy of the applied model. High
following:
R2 values (R2 > 0.80) indicate high proportion of variability explained
by the models and show good agreement between the calculated and WVP = 10.38 + 1.35X1 + 0.59X2 0.62X3 (13)
observed results within the range of experiment [20,21]. The coeffi-
cient of determination (R2) values for YI, WI, ΔE, UTS, SB, YM and WVP UTS = 10.98 1.64X1 + 1.33X3 0.78X1 X2 1.41X22 0.61X32 (14)
Table 2
Experimental data for the responses and films thickness.
a a
Run no. Responses Thickness (mm)
1 6.52 ± 0.32 114.51 ± 0.87 32.60 ± 0.37 23.62 ± 0.10 46.27 ± 0.06 62.95 ± 0.16 11.33 ± 0.65 0.13 ± 0.17
2 8.42 ± 0.79 119.40 ± 0.69 48.80 ± 0.96 23.53 ± 0.16 45.92 ± 0.10 64.04 ± 0.18 10.13 ± 0.87 0.10 ± 0.09
3 11.54 ± 0.52 85.00 ± 1.13 141.56 ± 0.83 18.09 ± 0.12 49.78 ± 0.12 51.02 ± 0.10 9.89 ± 0.49 0.11 ± 0.08
4 10.28 ± 0.28 96.00 ± 0.59 129.78 ± 0.72 17.93 ± 0.09 49.37 ± 0.16 50.84 ± 0.12 9.48 ± 0.57 0.12 ± 0.11
5 7.31 ± 0.85 129.00 ± 0.75 65.20 ± 0.68 27.91 ± 0.19 42.87 ± 0.09 77.80 ± 0.09 12.85 ± 0.76 0.15 ± 0.07
6 11.66 ± 0.52 101.60 ± 0.63 66.56 ± 1.13 19.11 ± 0.13 48.14 ± 0.11 55.13 ± 0.13 9.31 ± 1.10 0.10 ± 0.10
7 12.88 ± 0.45 93.00 ± 0.48 123.60 ± 0.20 15.80 ± 0.17 49.60 ± 0.15 44.33 ± 0.11 9.57 ± 0.93 0.10 ± 0.06
8 10.05 ± 0.78 89.00 ± 0.65 119.52 ± 0.53 17.37 ± 0.21 49.78 ± 0.07 48.42 ± 0.15 10.72 ± 0.47 0.11 ± 0.09
9 11.36 ± 0.98 84.00 ± 0.73 125.76 ± 0.59 18.89 ± 0.15 49.15 ± 0.13 51.33 ± 0.07 10.31 ± 0.66 0.10 ± 0.12
10 7.50 ± 0.23 124.50 ± 0.89 53.80 ± 0.39 18.76 ± 0.08 47.74 ± 0.17 52.17 ± 0.10 9.48 ± 0.83 0.11 ± 0.21
11 7.90 ± 0.24 102.00 ± 1.06 75.93 ± 1.24 16.76 ± 0.25 48.92 ± 0.13 46.79 ± 0.23 8.82 ± 0.73 0.09 ± 0.11
12 11.5 ± 0.36 89.98 ± 0.66 137.47 ± 1.44 18.68 ± 0.14 49.35 ± 0.06 51.72 ± 0.19 10.94 ± 0.91 0.11 ± 0.07
13 13.28 ± 0.36 92.71 ± 0.42 167.73 ± 0.63 13.11 ± 0.26 50.38 ± 0.19 37.98 ± 0.20 7.94 ± 1.06 0.09 ± 0.18
14 5.17 ± 0.84 127.96 ± 0.38 30.24 ± 0.30 25.76 ± 0.10 44.08 ± 0.11 68.89 ± 0.15 12.79 ± 0.52 0.12 ± 0.09
15 6.88 ± 0.58 87.92 ± 0.47 89.40 ± 0.87 19.82 ± 0.17 46.92 ± 0.05 55.13 ± 0.18 10.92 ± 0.38 0.10 ± 0.14
16 10.66 ± 0.37 114.00 ± 1.21 97.70 ± 0.56 16.63 ± 0.22 49.10 ± 0.10 46.79 ± 0.11 7.52 ± 0.89 0.08 ± 0.16
17 8.00 ± 0.69 119.00 ± 0.53 26.76 ± 0.69 26.60 ± 0.11 43.66 ± 0.16 72.68 ± 0.19 11.10 ± 1.19 0.13 ± 0.22
18 7.10 ± 0.39 105.00 ± 0.82 113.13 ± 0.43 21.58 ± 0.09 46.85 ± 0.14 59.11 ± 0.08 12.26 ± 0.69 0.12 ± 0.08
19 11.52 ± 0.44 96.00 ± 1.11 124.96 ± 0.89 18.68 ± 0.16 49.37 ± 0.15 51.72 ± 0.17 10.91 ± 0.70 0.10 ± 0.13
20 9.61 ± 0.55 87.92 ± 0.62 102.73 ± 0.75 17.72 ± 0.24 48.54 ± 0.18 49.17 ± 0.12 10.20 ± 0.71 0.11 ± 0.09
a
Mean ± S.D.
4
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Table 3
ANOVA study for the model fitting.
WVP × 10−8 (g/mhPa) UTS (Mpa) SAB (%) YM (Mpa) ΔE YI WI
R2,R2adj, PRESS and standard deviation (C.V %) defined for significant models (p˂0.01).
For parameters with two significant case, these defined for Quadratic model.
Table 4
Significance level (P) calculated for investigating the effect of gluten, CMC and CNF in to composite films.
Independent variable UTS (Mpa) SB (%) YM(Mpa) ΔE WI YI WVP × 10−8 (g/mhPa)
constant < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001**
Linear
X1 < 0.0001** < 0.0016** < 0.0136* < 0.0220* < 0.0001** < 0.0001** < 0.0001**
X2 < 0.002** < 0.9917 < 0.5569 < 0.5372 < 0.0147* < 0.0239 < 0.1673
X2 < 0.0001** < 0.0072** < 0.3655 < 0.6543 < 0.9374 < 0.7941 < 0.7086
Interaction
X1X2 < 0.1396 < 0.0031** < 0.0009** < 0.0003** < 0.0027** < 0.0035** < 0.0005**
X1X3 < 0.1452 < 0.2725 < 0.3672 < 0.0037** < 0.0229* < 0.0195* < 0.0280*
X2X3 < 0.3805 < 0.2433 < 0.0338* < 0.0679 < 0.1388 < 0.2486 < 0.1983
Quadratic
X12 < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001**
X22 < 0.0008** < 0.9772 < 0.0986 < 0.0445* < 0.0001** < 0.0002** < 0.0002**
X23 < 0.0005** < 0.0001** < 0.0179* < 0.1276 < 0.8510 < 0.5565 < 0.1715
Model
p-value Prob˃F < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001** < 0.0001**
F-value 19.05 33.25 24.40 32.83 75.66 69.05 80.40
SB = 90.03 + 9.92X1 + 3.33X3 + 7.20X1 X2 –3.79X2 X3 + 7.17X12 + 10.23X22 UTS values. Different reasons have been suggested for increasing of UTS
(15) by CNF: (1) formation of hydrogen bonds between CNF and biopolymer
which strengthens cross-linking between polymer chains (2) reducing
YM = 129.80 25.88X1 6.44X2 19.70X1 X2 13.81X1 X3 30.38X22 effect of CNF on the water content and thus decrease of water plasti-
18.08X32 (16) cizing effect (3) filling of void spaces between chains which decreases
chains local movement (4) developing of crystalline regions (5) higher
E = 18.27 + 4.06X1 + 1.14X2 + 0.89X1 X2 + 0.61X1 X3 + 0.8X12 + 1.04X22 mechanical strength of CNF in comparison to biopolymer matrix.
The linear and quadratic effects of gluten and quadratic effects of
+ 0.43X32 (17)
CNF concentrations and the interaction effects of gluten-CMC and CMC-
YI = 50.84 + 10.59X1 + 2.67X2 + 2.31X1 X2 + 1.70X1 X3 + 2.34X12 + 2.23X22 CNF concentrations on the strain at break (SB) value of the nano-
composite films were significant (P < 0.01). Increasing of gluten
+ 1.37X32 (18) concentration caused to increase of the strain at break (SAB) except at
low level of CMC which SAB value was high even at low concentration
WI = 49.46–2.07X1 –0.54X2 –0.64X1 X2 –0.32X1 X3 0.97X12 0.83X22 of gluten. The minimum amount of the SAB value could be observed at
0.65X32 (19) low level of gluten and high level of CMC. Although, the effect of CNF
on the SAB was significant (P < 0.05) however, it was not consider-
able.
3.2. Response surface Analysis
The linear and quadratic effects of gluten, quadratic effects of CMC
concentrations and the interaction effects of gluten-CMC and gluten-
3.2.1. Interaction effects of biopolymers and CNF on mechanical properties
CNF concentrations on the Young's modulus (YM) value of the nano-
In order to show the interactions effect between independent vari- composite films were significant (P < 0.01). The Young's modulus
ables (gluten, CMC and CNF concentrations) and the mechanical re-
amount of the composite films was increased by increasing of the gluten
sponses (ultimate tensile strength, strain at break (SB), Young's mod- concentration only in middle level of CMC concentration (similar trend
ulus), 3D response surfaces were used (Fig. 2). The linear and quadratic
with UTS) and middle to high levels of CNF. Increasing of CMC con-
effects of gluten and CNF concentrations on the UTS value of the na- centration increased YM value in all level of CNF.
nocomposite films were significant (P < 0.01) (Table 4). However,
CMC concentration did not show significant effect and only the inter-
action effect of gluten-CMC concentration was significant. The UTS 3.2.2. Interaction effects of biopolymers and CNF on water vapor
amount of the composite films was increased by increasing of the gluten permeability
concentration only in middle level of CMC concentration (Fig. 2a). In- In order to better explanation the relationships and interactions
creasing of CNF concentration caused to significant increase of UTS between independent variables and the WVP response, 3D response
value. Maximum amount of UTS could be observed at high level of surface plots as the function of gluten (X1), CMC (X2) and CNF (X3)
gluten and CNF and low level of CMC concentration. The increase of concentrations, were used (Fig. 3). The quadratic effects of gluten, CMC
gluten concentration could possibly increase the likelihood of dis- and CNF (X1, X2, X3) concentrations on the WVP value of the films were
ulphide bonds formation, which in turn results in the herein reported significant (P < 0.01) (Table 4). The increasing CNF concentration
5
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Fig. 2. Profile of response surface as Interaction effects of gluten, CMC and CNF concentration in nanocomposite films on (a) UTS (Mpa), (b) SB (%) and (c) YM
(Mpa).
reduced the WVP value of the composite films at all concentrations of cause to decrease of chains mobility which in turn decrease diffusion
gluten and CMC levels. While, increasing CMC concentration increased rate of water molecules and (5) lower permeability or hydrophilicity of
the WVP value at all levels of both gluten and CNF. The similar trend CNF in comparison to biopolymer matrix [24]. It seems that, the main
was observed for gluten as increasing gluten concentration caused to factor of reducing WVP in this research is filling of void spaces between
increasing of the WVP value at all levels of both CMC and CNF levels. biopolymer chains by CNF and decrease of chains mobility which in
The reducing effect of CNF on the WVP of the composite films, could be turn decrease diffusion rate of water molecules.
attributed to the following reasons: (1) formation of hydrogen bond by The effect of gluten, CMC and their interaction on the WVP of the
a dipole–dipole force between the matrix and the hydroxyl group of composite films can be explained by different mechanisms. According
nano cellulose and therefore improving the matrix cohesiveness (2) the SEM images, it can be seen that, the addition of gluten in the matrix
formation of hydrogen bonds between CNF and biopolymer matrix also of carboxymethyl cellulose film and their interaction together, creates
may lead to decrease in free hydrophill groups (OH), (3) formation of pores and porous networks in the form of a sponge structure. Therefore,
long and zigzag pathway thorough the film thickness against pene- it seems that the increasing the concentration of gluten and carbox-
trating molecules, (4) filling of void spaces between biopolymer chains ymethyl cellulose and their interactions with together, increases the
6
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Fig. 3. Profile of response surface as Interaction effects of gluten, CMC and CNF concentration in nanocomposite films on (a) WVP.
amount of pores and cavities created in the film matrix and thus in- maximum desirability, the desired goal for three variable concentra-
creases the WVP of the composite film. tions were set within the range, while the UTS, SAB, ΔE, YM and WI
were set to maximum and WVP was defined as minimum. The most
3.2.3. Interaction effects of biopolymers and nanoparticle on color important value of 5, indicated by (+++++), was chosen only for
properties UTS and the important value of 4 (++++), was chosen for SB, YM
Packaging film color is one of the important indicator and effective and WVP. Also, the important value of 3 (+++), was chosen for WI, YI
in consumer acceptance. In this study, in order to evaluate the color and ΔE index. The optimum concentration levels of gluten (% w/w),
properties of the films, the color difference (ΔE), yellowness index (YI) CMC (% w/w) and CNF (% w/w) were detected as 1 (wt%), 0.686 (wt
and whiteness index (WI) of the films were investigated. %) and 8.549 (%), respectively with maximum desirability 87%. The
The linear and quadratic effect of gluten concentration on all color maximum desirability value (0.878) means that this design is suitable
parameters were significant (P < 0.01) (Table 4). However, similar for using.
trend for CMC concentration on color parameters not be seen and CMC
concentration on color parameters had not significant. Also, CNF con- 3.4. Verification experiments and validation of the model equations
centration had not significant effect on any color indexes. The inter-
action effect of gluten-CMC and gluten-CNF were significant for all In order to validate the adequacy of the model equations, a ver-
color parameters (P < 0.05). ification experiments were performed with optimal variable con-
To perception and visualize the combined effect of independent centrations gluten 1 (wt%), CMC 0.686 (wt%) and CNF 8.549 (wt%
variables (gluten, CMC and CNF) on ΔE, WI and YI amounts of the films, based on gluten and CMC biopolymers solid matter in each formula).
3D response surfaces plots have been presented (Fig. 4). The color The experimental and predicted data for the responses at optimum
difference (ΔE) increased by enhancing of gluten concentrations at all point are shown in Table 6.
levels of CMC and CNF however, this effect was more strength at higher Only a small percentage error was notified between the experi-
concentration of CMC. The maximum ΔE could be observed at max- mental and predicted values and the results from the experimental re-
imum amount of gluten, CMC and CNF. This showed gluten had un- sponse values were reasonably close to the predicted values. Thus,
desirable influence on overall color change of the films. verification experiments demonstrated the adequacy of the response
Similar trend was observed for YI value. The WI value of the com- surface equations for responses.
posite films was decreased by increasing of gluten concentration at all
concentrations of CMC and CNF. The maximum WI value could be seen 3.5. Morphological properties of films
in the low concentration of gluten and middle to high concentration of
CMC and CNF. 3.5.1. Field emission scanning electron microscopy (FE-SEM)
The morphology of the cross-section and surfaces of the WG
3.3. General optimization of variables for WG/CMC/CNF film (1.75 wt%)/CMC (0.75 wt%), WG (1.75 wt%)/CMC (0.75 wt%)/CNF
5% and WG (1.75 wt%)/CMC (0.75 wt%)/CNF 10% nanocomposite
Fitted regression models obtained from the statistical analysis by films were studied with FE-SEM (Fig. 5). According on the provided
RSM led tofinding the optimum concentration of gluten, CMC and CNF images, the WG/CMC composite films without CNF (0%) showed a
according to the WVP, mechanical and color responses of films sepa- relatively uniform and smooth surface. However, the cross-section
rately, as shown in Table 5. In RSM, the desirability function (DF) is images of WG/CMC film showed a loose and porous structure with a lot
widely used to determine a combination of variables to optimize mul- of holes and cracks. The FE-SEM images showed that adding CNF (5%)
tiple responses and provide the most desirable responses (global de- to the biopolymer matrix lead to more heterogeneous structure and
sirability). With define the desired goal for each factors and response, irregularity with rough surface. The cross section image showed that
software suggests one solution with the highest desirability. The scale of adding of CNF decreased the pores and cavities in the nanocomposite
the desirability function ranges between 0 (completely undesirable re- structure. With increasing CNF concentration to 10%, the spherical
sponse) and 1(fully desired response) and it corresponds to the desir- particles were observed at the surface of the film, however, cross sec-
ability interpretation such as: very good (1.0–0.8); good (0.8–0.63); fair tion images indicated relatively more compact and coherent structure
(0.63–0.37); poor (0.37–0.2); and very poor (0.2–0). To obtain with lowe voids and pin holes.
7
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Fig. 4. Profile of response surface as Interaction effects of gluten, CMC and CNF concentration in nanocomposite films on (a) ΔE, (b) YI and (c) WI.
Table 5
Optimization values of gluten/CMC/CNF nanocomposite films for each category parameters separately.
Factor name Goal Optimized value Factor name Goal Optimized value Factor name Goal Optimized value
gluten (%) In range 1.000 gluten (%) In range 1.000 gluten (%) In range 1.200
CMC (%) In range 0.685 CMC (%) In range 0.816 CMC (%) In range 0.542
CNF (%) In range 9.260 CNF (%) In range 8.080 CNF (%) In range 7.089
UTS (MPa) Max 13.28 ΔE Min 12.949 WVP × 10−8 (g/mhPa) Min 7.114
SAB (%) Max 115.476 YI Min 38.022 Desirability 100%
YM (Mpa) Max 128.092 WI Max 50.38
Desirability 80% Desirability 100%
8
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Fig. 5. FE-SEM images of (a, b) WG/CMC, (c, d) WG/CMC/CNF 5%, (e, f) WG/CMC/CNF 10% nanocomposite films.
9
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
Fig. 6. AFM images of (a) WG/CMC, (b) WG/CMC/CNF 5%, (c) WG/CMC/CNF 10%, nanocomposite films.
Fig. 7. XRD patterns of (a) WG/CMC, (b) WG/CMC/CNF 5%, (c) WG/CMC/CNF 10% and (d) pure CNF films.
nanoparticles did not cause to development of crystalline regions in the increasing CNF concentration. On the other hand, there was no high
biopolymer, which means that the CNF nanoparticles did not act as new intensity crystalline peak in XRD patterns which could be attributed to
nuclei to create crystalline areas. the proper distribution of CNF in matrix and amorphous structure of the
resulted films.
4. Conclusion
Acknowledgement
In this research work, biodegradable materials (WG, CMC and CNF)
were used for producing of new nanocomposite film and optimization The authors gratefully acknowledge the support of food biophysics
of the film formulation was carried out by RSM method. The RSM laboratory of Tabriz University and Nutrition Research Center of Tabriz
models were developed for seven responses including mechanical, color University of Medical Sciences.
and WVP features of the resulted films. The results from the verification
experimental response values were in good agreement with the pre- References
dicted values. Verification experiments demonstrated the adequacy of
the response surface equations for responses. The FE-SEM images [1] Ali. Abdulkhani, Maryam. Daliri Sousefi, Alireza. Ashori, Ghanbar. Ebrahimi,
Preparation and characterization of sodium carboxymethyl cellulose/silk fibroin/
showed that the addition of the optimum level of the CNF in composite graphene oxide nanocomposite films, Journal of Polymer Testing 52 (2016)
films matrix can improve the cohesion of the films. Also, the results of 218–224.
the AFM analysis indicated that the surface roughness increased with [2] B. Ghanbarzadeh, A.R. Oromiehi, Biodegradable biocomposite films based on whey
10
V. Bagheri, et al. Polymer Testing 78 (2019) 105989
protein and zein: barrier, mechanical properties and AFM analysis, International food packaging, Journal of Critical Reviews in Food Science and Nutrition 00-00
Journal of Biological Macromolecules 43 (2) (2008) 209–215. (2016).
[3] Hui Yin. Fan, Daniel. Duquette, Marie-Josée. Dumont, Benjamin K. Simpson, [14] J. Zhao, X. He, Y. Wang, W. Zhang, X. Zhang, X. Zhang, Y. Deng, C. Lu,
Salmon skin gelatin-corn zein composite films produced via crosslinking with glu- Reinforcement of all-cellulose nanocomposite films using native cellulose nanofi-
taraldehyde: optimization using response surface methodology and characteriza- brils, Journal of Carbohydrate Polymers 104 (2014) 143–150.
tion, International Journal of Biological Macromolecules 120 (2018) 263-27. [15] N.A. El-Wakil, E.A. Hassan, R.E. Abou-Zeid, A. Dufresne, Development of wheat
[4] Y. Song, Q. Zheng, C. Liu, Green bio composites from wheat gluten and hydro- gluten/nanocellulose/titanium dioxide nanocomposites for active food packaging,
xyethyl cellulose: Processing and properties, Journal of Industrial Crops and Journal of Carbohydrate Polymers 124 (2015) 337–346.
Products 28 (2008) 56–62. [16] Tanara. Sartori, Gabriela. Feltre, Paulo. Jose do Amaral Sobral, Rosiane. Lopes da
[5] S. Tunc, H. Angellier, Y. Cahyana, P. Chalier, N. Gontard, E. Gastaldi, Functional Cunha, Florencia. Cecilia Menegalli, Properties of films produced from blends of
properties of wheat gluten/montmorillonite nanocomposite films processed by pectin and gluten, Journal of Food Packaging and Shelf Life 18 (2018) 221–229.
casting, Journal of Membrane Science 289 (2007) 159–168. [17] M.A. Bezerra, R.E. Santelli, E.P. Oliveira, L.S. Villar, L.A. Escaleira, Response sur-
[6] F. Rafieian, M. Shahedi, J. Keramat, J. Simonsen, Mechanical: thermal and barrier face methodology (RSM) as a tool for optimization in analytical chemistry, Journal
properties of nano-bio composite based on gluten and carboxylated cellulose na- of Talanta 76 (2008) 965–977.
nocrystals, Journal of Industrial Crops and Products 53 (2014) 282–288. [18] B.S. Kayserilioğlu, U. Bakir, L. Yilmaz, N. Akkas, Drying temperature and relative
[7] A. Duval, S. Molina-Boisseau, C. Chirat, Comparison of Kraft lignin and lingo sul- humidity effects on wheat gluten film properties, Journal of Agricultural and Food
fonates addition to wheat gluten-based materials: mechanical and thermal prop- Chemistry 51 (2003) 964–968.
erties, Journal of Industrial Crops and Products 49 (2013) 66–74. [19] L. Nouri, A. Mohammadi-Nafchi, Antibacterial, mechanical, and barrier properties
[8] N. Gontard, C. Duchez, J.L. Cuq, S. Guilbert, Edible composite films of wheat gluten of sago starch film incorporated with betel leaves extract, International Journal of
and lipids: water vapor permeability and other physical properties, International Biological Macromolecules 66 (2014) 254–259.
Journal of Food Science & Technology 29 (1994) 39–50. [20] M. Dak, M.K. Jain, S.L. Jat, Optimization of microwave-vacuum drying of pome-
[9] H. Almasi, B. Ghanbarzadeh, A.A. Entezami, Physicochemical properties of granate arils, Journal of Food Measurement and Characterization 8 (4) (2014)
starch–CMC–nano clay biodegradable films, International Journal of Biological 398–411.
Macromolecules 46 (2010) 1–5. [21] D.S.A. Delfiya, K. Thangavel, D. Amirtham, Preparation of curcumin loaded egg
[10] P. Rachtanapun, S. Luangkami, K. Tanprasert, R. Suriyatem, Car-boxymethyl cel- albumin nanoparticles using acetone and optimization of desolvation process, The
lulose film from durian rind, Journal of LWT—Food Science and Technology 48 Protein Journal 35 (2) (2016) 124–135.
(2012) 52–58. [23] R. Rodrigo, C.A. Toro, J. Cuellar, Influence of the geometric factors of the experi-
[11] H.M.C. Azeredo, K.W. Waldron, Crosslinking in polysaccharide and protein films mental device used in suspension polymerization on the properties of poly (styrene-
and coatings for food contact – a review, Journal of Trends in Food Science & co-divinyl benzene) micro particles, Journal of Applied Polymer Science 124 (2)
Technology 52 (2016) 109–122. (2012) 1431–1446.
[12] F.M. Pelissari, P.J. do Amaral Sobral, F.C. Menegalli, Isolation and characterization [24] N. Noshirvani, B. Ghanbarzadeh, H. Fasihi, H. Almasi, Starch–PVA nanocomposite
of cellulose nanofibers from banana peels, Journal of Cellulose 21 (2014) 417–432. film incorporated with cellulose nanocrystals and MMT: a comparative study,
[13] F. Vilarinho, A. Sanches-Silva, M. Vaz, J. Farinha, Nanocellulose: a benefit for green International Journal of Food Engineering 12 (2016) 37–48.
11