Food Chemistry 151 (2014) 343351

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Biocomposite cellulose-alginate lms: Promising packaging materials

Juho Antti Sirvi a, Aleksi Kolehmainen a, Henrikki Liimatainen b, Jouko Niinimki b, Osmo E.O. Hormi a,
a
b

Department of Chemistry, University of Oulu, P.O. Box 3000, FI-90014, Finland

Fibre and Particle Engineering Laboratory, University of Oulu, P.O. Box 4300, FI-90014, Finland

a r t i c l e

i n f o

Article history:
Received 19 April 2013
Received in revised form 19 September 2013
Accepted 6 November 2013
Available online 16 November 2013
Keywords:
Cellulose
Nanocellulose
Reinforcement
Alginate
Biocomposite
Ionic cross-linking

a b s t r a c t
Biocomposite lms based on cellulose and alginate were produced using unmodied birch pulp,
microbrillated cellulose (MFC), nanobrillated cellulose (NFC) and birch pulp derivate, nanobrillated
anionic dicarboxylic acid cellulose (DCC), having widths of bres ranging from 19.0 lm to 25 nm as cellulose bre materials. Ionically cross-linked biocomposites were produced using Ca2+ cross-linking. Addition of micro- and nanocelluloses as a reinforcement increased the mechanical properties of the alginate
lms remarkably, e.g. addition of 15% of NFC increased a tensile strength of the lm from 70.02 to
97.97 MPa. After ionic cross-linking, the tensile strength of the lm containing 10% of DCC was increased
from 69.63 to 125.31 MPa. The biocomposite lms showed excellent grease barrier properties and
reduced water vapour permeability (WVP) after the addition of cellulose bres, except when unmodied
birch pulp was used.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Composite materials derived from biobased sources have drawn
lots of attention during the recent years (Hansen & Plackett, 2008).
One of the most studied applications of biocomposites is in environmentally friendly food packaging materials (Kristo, Biliaderis, &
Zampraka, 2007). Natural polymers, such as polysaccharides, are
ideal starting materials for these kinds of composites due to their
biodegradability, biocompatibility, low toxicity and renewability.
Cellulose is a linear polysaccharide of (14)-linked b-D-glucose,
produced by a variety of plants and algae, along with some bacteria. In nature, cellulose is produced as brous material which has
long been used as a main component in paper, and as a starting
material e.g. for edible packages (Ayranci & Tunc, 2001; Cheng,
Abd Karim, & Seow, 2008). Additionally, the use of cellulose microand nanobres in composite materials has been a scientic interest
in recent years (Abdul Khalil, Bhat, & Ireana Yusra, 2012). Cellulose
micro- and nanobres have been studied as reinforcement agents
in synthetic (Zoppe, Peresin, Habibi, Venditti, & Rojas, 2009) and
natural polymers (Johnson, Zink-Sharp, Renneckar, & Glasser,
2009).
Alginate is another type of linear polysaccharide, consisting of
(14)-linked b-D-mannuronate and a-L-guluronate units. The relative proportions of mannuronate and gulunorate vary within the
source of the alginate (Johnson, Craig, & Mercer, 1997). The most
Corresponding author. Tel.: +358 294 48 1631; fax: +358 8 553 1593.
E-mail address: osmo.hormi@oulu. (O.E.O. Hormi).
0308-8146/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2013.11.037

common source of alginate is the cell wall of brown algae, and

for commercial purposes alginate is extracted from seaweed. Alginate and alginate-based biocomposites have been used as, for
example, adsorbents for ionic dyes (Sui et al., 2012), antibacterial
lms (Benavides, Villalobos-Carvajal, & Reyes, 2012), wound healing materials (Sikareepaisan, Ruktanonchai, & Supaphol, 2011), and
stimulus response drug releasing materials (Mourio et al., 2011).
Although alginate based biocomposites have strong potential to
be used for example in packaging applications, they usually lack
the desired mechanical properties, such as high strength. To
increase the mechanical properties of alginate based biocomposites, cellulose nanocrystals have been studied as reinforcement
agents in alginate lms (Cheng, Lu, Zhang, Shi, & Cao, 2012). However, there is only a scarcity of information on the mechanical
properties of biocomposite materials with an alginate matrix reinforced with different types of wood cellulose bres. In this paper,
four different wood cellulose bres, with widths ranging from
19.0 lm to 25 nm, were used as reinforcement agents for alginate
biocomposites lms to be utilised in application such as packaging.
Effects of the different bres on the lm properties and on the
compatibility between cellulose bres and alginate were studied.
Moreover, an ionic cross-linking of the biocomposite lms with
Ca2+ ion, and grease and water vapour barrier properties were also
studied. Properties of lms, and the compatibility between cellulose bres and alginate were studied by Fourier transfer infrared
spectroscopy (FTIR), eld emission scanning electron microscopy
(FESEM), energy-dispersive X-ray spectroscopy (EDS) and wideangle X-ray diffraction (WAXD).

344

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

2. Experimental section
2.1. Materials
Bleached birch pulp (Betula verrucosa and pendula) was obtained as dry sheets and used as a bre material after being disintegrated in deionised water. The properties of cellulose are
presented elsewhere (Liimatainen, Sirvi, Haapala, Hormi, & Niinimki, 2011). The commercial grade microbrillated wood cellulose (MFC, KY-100G) was obtained from Daicel Industries Ltd.
(Japan).
The combined treatment of Valley beating and high-pressure
homogenisation was used to prepare nanobrillated celluloses
(NFC) from the birch pulp. Prior to homogenising, the bleached
birch was ground in a Valley beater to a dry weight of 2.0% over
50 h, yielding 5503 g of weight. The samples where then diluted
to a dry weight of 0.7% and homogenised with a homogeniser
(Invensys APV 2000, Denmark) at a pressure of 3501200 bar to
create NFC. NFC had same charge density as original pulp
(0.08 mmol/g). Charge density is mainly due to the presence of
hemicelluloses.
Nanobrillated dicarboxylic acid cellulose (DCC) with a charge
density of 0.75 mmol/g was prepared as reported earlier
(Liimatainen, Visanko, Sirvio, Hormi, & Niinimaki, 2012).
Sodium alginate (Scogin XL, low viscosity) was kindly donated
by FMC Biopolymer (Ireland). Oil Red O (1-([4-(xylylazo)
xylyl]azo)-2-naphthol), used in grease barrier tests, was obtained
from SigmaAldrich (Germany) and Ottawa-grade sand (particle
size 2030 mesh) from Fisher Scientic (UK) and Turpentine (Oulu,
A1 Wood turpentine, Finland) from Kiilto Ltd. All of other
chemicals used in the periodate and chlorite oxidations (NaIO4,
NaClO2 and CH3COOH), aldehyde and carboxyl content analyses
(NH2OHHCl, CH3COONa2H2O, NaCl, and NaOH) and ionical
crosslinking (CaCl22H2O) were obtained as p.a. grade from
SigmaAldrich (Germany) and were used without further purication. Deionised water was used throughout the experiments.
2.2. Preparation of biocomposite lms
The biocomposite lms based on cellulose and alginate were
prepared using the solvent-casting method. At rst, alginate was
slowly dissolved in deionised water. Then cellulose (unmodied
birch pulp, MFC, NFC or DCC) dispersion was added to this solution,
whilst varying the amount of cellulose from 0% to 50% of the lm
dry weight. Total volume of the alginate and cellulose dispersions
was 135 ml and a dry weight of lms was 1.515 g, resulting in lms
with a grammage of 30 g/m2. After addition of the cellulose dispersion, the resulting alginate/cellulose mixture was dispersed using
an UltraTurrax disperser (Germany) (11,000 rpm) for ten minutes
to obtain a homogeneous dispersion after which a water circulation vacuum was used to remove air bubbles from the dispersion.
Finally, the dispersion was cast in a polystyrene tray of 500 cm2
area. Films were allowed to dry from 3 to 5 days at room temperature to obtain 13 biocomposite lms with 1%, 10%, 15%, 20% and
50% of MFC, 1%, 5%, 10% and 15% of NFC, 1% and 10% of DCC, and
50% of birch pulp of the lm dry weight were, designated as
MFC1, MFC5, MFC10, MFC15, MFC20, MFC50, NFC1, NFC5, NFC10,
NFC15, DCC1, DCC10 and BIRCH50, respectively. Pristine alginate
lm was also produced as a reference.
2.3. Preparation of ionically cross-linked biocomposite lms
In order to ionically cross-link a lm, it was immersed in 400 ml
of 10% CaCl22H2O solution for 30 min. The lm was then the carefully removed from the solution and rinsed thoroughly with

deionised water. The lm was placed on a polycarbonate plate

and small weights were placed on the edges of the lm to prevent
the shrinkage of the lm during air-drying (24 h). Ionic crosslinking was conducted with a pristine alginate, BIRCH50, MFC50,
NFC10 and DCC10 lms and they were designated as ICBICH50,
ICMFC50, ICMFC10, ICDCC10, respectively.
2.4. Mechanical properties
The lms were cut into 15  150 mm stripes and conditioned at
23 C and 50% humidity for 48 h before testing. Specic tensile
strength (r/q), elongation at break and specic modulus (E/q)
were determined according to ISO Standard, 1995 and Lorentzen
& Wettre tensile strength tester (Sweden). A test speed of
50 mm/min was used. Initial distance between grips was set to
100 mm. Each test was replicated ve times and the results are
presented as averages of these measurements. Tensile strength
(r) and modulus (E) was calculated using the density (q) from r/
q and E/q, respectively.
Film thicknesses were determined as averages of ve random
positions of the lm using Lorentzen & Wettre thickness metre
(Sweden).
2.5. Field emission scanning electron microscope and energydispersive X-ray spectroscopy
FESEM (Zeiss Ultra Plus, Carl Zeiss SMT AG, Germany) was used
to examine morphologies of used cellulosics and the surface of the
lms. Fibres were sputter-coated with Pd, and an accelerating voltage of 5 kV was used whilst the lms were sputter-coated with carbon, and an accelerating voltage of 15 kV was used. EDS was used
to obtain surface elemental analysis of the lms.
2.6. Fourier transform infrared spectrum
FTIR spectra of lms were recorded using a Bruker FT-IR spectrometer (USA) using spectral width ranging from 4000 to
400 cm1 with 2 cm1 resolution and an accumulation of 32 scans.
2.7. X-ray diffraction
For X-ray diffraction (XRD) analyses, samples were analyzed between 2h = 5 and 114 with angle step size 2h = 0.02 in a D8 Discover X-ray Diffractometer (Bruker AXS Inc., Madison, MI, USA)
equipped with a Co Ka (40 kV/35 mA) source. Pure birch pulp,
MFC, NFC and DCC lms were prepared by casting a 1% solution
of cellulose dispersion on a Teon plate to obtain a lm with a
grammage of 30 g/m2. The degree of crystallinity of pure cellulose
lms in terms of the crystallinity index (CrI) was calculated from
the peak intensity of the main crystalline plane (0 0 2) diffraction
(I002) at 22.8 and from the peak intensity at 18.0 associated with
the amorphous fraction of cellulose (Iam) according to following
equation (Segal, Creely, Martin, & Conrad, 1959)

CrI


I002  Iam
 100%
I002

2.8. Grease barrier properties

The grease resistance of the lms was determined according to
TAPPI Standard, 2010. Each test was replicated four times: two
times for the bottom layers and two times for the upper layers of
the 50  50 mm2 lm specimens.
The method involves the use of turpentine colored with an oilsoluble red dye to measure the penetration of the grease through a

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

greaseproof material. Moisture-free colored turpentine was

prepared by mixing 100 ml of turpentine with 1.0 g of Oil Red O
red dye and 5 g of anhydrous CaCl2 in a ask. The ask was closed
with stopper, and the mixture was shaken well and allowed to
stand for 10 h. Finally, the mixture was ltrated and stored in an
airtight bottle.
For grease barrier tests, a lm was placed on a white printing
paper, and 5 g of sand was placed on top of the lm. A volume of
1.1 ml of colored turpentine was added to the sand pile. The lm
was carefully slid to a new unexposed portion of the paper, and
the time when colored-turpentine-stained paper was observed
was recorded. During the rst minute, the lm was slid every
15 s, then at 1 min intervals for the next 5 min and nally at
5 min intervals until a maximum time of 30 min was reached.
2.9. Water vapour permeability
Water vapour permeability (WVP) was determined according to
ASTM Standard, 1995. All lms were conditioned at 23 C and 50%
humidity for 48 h before being tested. Schott Duran laboratory
glass bottles with a hole drilled in their caps were lled to onefourth their volume with deionised water. The test specimen was
placed between the bottle and cap, and a rubber seal was used to ensure a tight seal. The test assembly was weighed and placed in a condition-controlled chamber (23 C and 50% humidity, equipped with
a fan to ensure good circulation of air inside the chamber) and the
weight lost was recorded in one hour intervals for 8 h. A steady state
was reached with all samples during the rst hour. The slope of the
strain line of weight lost vs. time was calculated, and the water vapour transmission rate (WVTR) was determined by dividing the
slope by the exposed lm area. WVP was calculated as follows:

WVP

WVTR
SR1  R2

where S is the saturation vapour pressure at the test temperature

(28  102 Pa at 23 C), R1 is the relative humidity in the dish expressed as a fraction (1), and R2 is the relative humidity at the vapour sink expressed as a fraction (0.5).
3. Results and discussion
3.1. Cellulose and alginate biocomposite lms
3.1.1. Preparation of biocomposite lms
FESEM images of the cellulose bres used to obtain biocomposite lms are presented in Supplementary data (Fig. S1). The untreated birch pulp consisted mainly of cellulose bres (width
19.0 lm) with 3.4% of nes, according to FibreMaster image analyzer. MFC consisted of long brils, but also included larger micro-sized particles such as pieces of cell walls. NFC consisted
mainly of nanobril bundles with widths smaller than 100 nm,
whilst DCC contained single nanobrils and nanobril bundles
with widths of below 25 nm. However, it is possible that the brils
existing in the suspension are actually thinner compared to the brils observed in FESEM images due to the agglomeration of nanobres during the sample preparation (Liimatainen et al., 2012).
The appearance of dried lms depended on the amount and the
type of cellulose used in the lm. Alginate lms with no added cellulose were fully transparent, and the transparency decreased as
the amount of cellulose increased: MFC50 and BIRCH50 were turbid with low transparency. Due to the small bre sizes, lms with
NFC and DCC were only slightly less transparent compare to pristine alginate lm. All the lms were exible and could be easily
handled. Films with cellulose had higher apparent roughness than
the pure alginate lm. The surface facing towards the tray during

345

the drying was signicantly smoother compared to the surface

facing towards the air during the drying. Nano-sized celluloses
(NFC and DCC) provided smoothest surfaces, whereas MFC and
especially unmodied birch pulp reinforced lms were signicantly rougher. The visual appearance of the lms obtained from
FESEM is shown in a later chapter.
3.1.2. Preparation of ionically cross-linked biocomposite lms
Cross-linking can be used to increase the strength of materials
(Bhattacharya, Rawlins, & Ray, 2009). It can also be used to
decrease water solubility of polymers, such as polysaccharides
(Ramesh, Hasegawa, Sugimoto, Maki, & Ueda, 2008). To obtain
cross-linked biocomposite lms, Ca2+ ions were used as an ionic
crosslinker. The mechanism of the ionic cross-linking of alginate
is well described in literature (Sikorski, Mo, Skjk-Brk, & Stokke,
2007). During the drying, fractures were observed in the pristine
alginate lm and intact ionically cross-linked alginate lm could
not be produced. Fractures may be due to the high tension caused
by lm shrinking. However, intact ionically cross-linked lms from
BIRCH50, MFC50, NFC10 and DCC10 were obtained.
The appearances of ICBIRCH50 and ICMFC50 were smooth and
they were only slightly wrinkled on the edges. ICNFC10 in turn
was more wrinkled and bent, whilst ICDCC10 was smooth and only
slightly bent on the edges of the lm. Minor fractures were observed at the edges of ICDCC10. Due to the higher amount of alginate, ICNFC10 and ICDCC10 were more prone to shrinking, and
hence they were more wrinkled compared to ICBIRCH50 and
ICMFC50.
All lms dissolved or disintegrated in water without cross-linking treatment. After cross-linking the lms still wetted easily, but
they did not dissolve in water. It is known that cross-linking of
alginate with multivalent ions decreases dramatically the water
solubility of the alginate (Nakamura, Nishimura, Hatakeyama, &
Hatakeyama, 1995).
3.2. Fourier transform infrared spectrum
FTIR spectra of the pristine alginate lm, MFC50, NFC10, DCC10
and BIRCH50 before and after crosslinking are presented in Fig. 1
and assignments of the most characteristic vibration modes of all
of the biocomposite lms determined by literature (Fuks, Filipiuk,
& Majdan, 2006; Papageorgiou et al., 2010) are presented in
Supplementary data (Table S1). In the spectrum of the pristine alginate lm, characteristic vibration bands of m(COO)asym and
m(COO)sym are observed at 1613.66 and 1414.54 cm1, respectively,
and the band of CO stretching vibrations is observed at
1038.02 cm1.
In the FTIR spectra of the biocomposite lms m(COO)asym peak is
shifted to a lower wavenumber as the amount of cellulose is increased. The band of m(COO)asym of MFC1 is at 1614.42 cm1 and
it is shifted to 1607.97 cm1 when the amount of MFC is increased
to 50%. The lowest m(COO)asym wavenumber is observed with
BIRCH50 (1603.19 cm1). This may be due to the presence of large
amount of hemicelluloses in birch pulp, which may have better
interaction with alginate than cellulose. The CO stretching vibration bands in MFC50, BIRCH50, and NFC10 are also shifted to
slightly lower wavenumber and are at 1031.51, 1027.27 and
1026.18 cm1, respectively. Wavenumber of the peak of CO of
DCC10 is close to pristine alginate lm, 1035.50 cm1, which indicates that, compared to other cellulose materials, anionic DCC has
lowest interaction with alginate.
After the crosslinking, the m(COO)asym band of the pristine alginate lm is shifted to a lower wavenumber (1602.78 cm1),
whereas the m(COO)sym band has a slightly higher wavenumber
(1420.94 cm1). This is due to the different interactions between
monovalent sodium and divalent calcium ions with alginate

346

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

Fig. 1. FTIR spectra of alginate and biocomposite lms (a) before and (b) after cross-linking.

(Papageorgiou et al., 2010). From the Table 1 it can be observed

that Dm(COO) (m(COO)asym  m(COO)sym) of the pristine alginate is
decreased from 199.12 to 181.84 cm1 after ionic cross-linking,
which indicates that a more uniform, structured, carboxylate
groups are formed during the cross-linking (Fuks et al., 2006). It
has been proposed that there is pseudo bridged unidentate
complex between alginate and multivalent metal cations
(Papageorgiou et al., 2010). However, recent density functional
calculations have shown that there is only an ionic interaction
between alkaline metals and alginate (Agulhon, Markova, Robitzer,
Quinard, & Mineva, 2012).
Values of Dm(COO)Ca of the ICMFC50 and ICBIRCH50 have also
been shifted from 194.12 to 181.01 cm1 and from 188.90 to
175.12 cm1, respectively. However, values of Dm(COO) of lms
containing nano-sized cellulose are virtually the same before and
after cross-linking. Nano-sized celluloses may interact with calcium alginate which leads to the change of the bonding between
alginate and the calcium ion.
It has been previously proposed that anionically charged
nanocelluloses can take part in the formation of highly cross-linked
alginate after calcium cross-linking (Lin, Bruzzese, & Dufresne,
2012). Nanocelluloses used in the present study had anionic
charges of 0.08 mmol/g (NFC) and 0.75 mmol/g (DCC) and thus
NFC and DCC may form ionic cross-linking network with calcium
alginate. Even though birch pulp has same anionic charge as NFC,
it has signicantly larger bre size and lower surface area, due to
which birch pulp has fewer ionic groups available to interact with
calcium alginate than NFC.
3.3. Field emission scanning electron microscope and energydispersive X-ray spectroscopy
FESEM images of biocomposite lms are presented in the Fig. 2.
Images are from the surface faced towards the tray during the drying. The lms were observed to be very sensitive and especially the

Table 1
Asymmetric and symmetric carboxyl vibration frequencies of alginate.

lms with high amount of alginate were easily destroyed during

the zooming and thus only large scale images were obtained. The
pristine alginate lm (Fig. 2a) has a smooth and even surface
(white spots on the surface are holes formed on the surface due
to the electron beams of the microscope). The surface of MFC50
(Fig. 2b) was also relative smooth and even, however, some cellulose brils are buried under alginate matrix. Compared to MFC50,
the surface of BIRCH50 (Fig. 2c) is signicantly less even and large
pores can be observed on the surface. This is due to the much larger
bre size of birch pulp compared to MFC. Due to the nano-sized bres, the surfaces of NFC10 (Fig. 2d) and DCC10 (Fig. 2e) are
smooth and no individual bres can be observed in the FESEM
images.
After ionic cross-linking, the surfaces of pristine alginate, NFC10
and DCC10 lms were smooth and identical to non-cross-linked
lms (images not shown). However, the surface of ICBIRCH50
(Fig. 2f) was much more even and fewer pores were observed compared to BIRCH50. It is possible that shrinkage of alginate during
the cross-linking compresses the BIRCH50 lm, which in turn reduces the pore sizes of the lm. ICBIRCH50 (Fig. 2g) surface faced
towards the air during the drying was also examined and it is observed to be rougher having individual cellulose bres clearly visible. In a comparison, the surface of ICMFC50 (Fig. 2h) was also
observed to be rough, but due to the smaller bre size, no individual bres or visible pores can be observed on its surface.
Results from EDS-elemental analysis are presented in Supplementary data (Table S2). All surface sodium ions were changed
to calcium ions during the cross-linking. In case of ICNFC10, 0.6%
of chlorine is still present in the surface, which indicates that small
amount of unreacted CaCl2 remains on the surface of lm after
washing. Interestingly, all lms except ICMFC50 have similar
amounts of calcium ions on their surfaces (17.020.0%). ICMFC50,
which has only 3.9% of calcium in its surface, may have higher
amount of cellulose bres on the surface compared to other
biocomposite lms, which reduces the amount of calcium bound
to the surface of lm.
3.4. Wide-angle X-ray diffraction

Sample

m(COO)asym

m(COO)sym

Dm(COO)

Alg
MFC50
NFC10
DCC10
BIRCH50
ICAlg
ICMFC50
ICNFC10
ICDCC10
ICBIRCH50

1613.66
1607.97
1609.31
1609.13
1603.19
1602.78
1606.32
1609.31
1614.54
1598.18

1414.54
1413.85
1420.79
1414.78
1414.29
1420.94
1425.31
1420.79
1421.09
1422.48

199.12
194.12
188.52
194.35
188.9
181.84
181.01
188.52
193.45
175.7

X-ray diffractograms of the pristine alginate, BIRCH50 MFC50,

NFC10 and DCC10 lms before and after ionic cross-linking are
presented in Fig. 3, along with the diffractograms of the pure birch
pulp, MFC, NFC, DCC lms. All celluloses showed characteristic
peaks of cellulose I with the main 2h diffraction angles close to
14.5, 16.0, and 22.2 associated with the 101, 101 and 0 0 2 crystalline planes, respectively. CrIs of the birch pulp, MFC, NFC and DCC
were 56.6%, 63.1%, 48.8% and 44.9%, respectively, which are consistent with the literature (Iwamoto, Nakagaito, & Yano, 2007). On the
other hand, the diffractogram of the pristine alginate lm shows

347

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

Fig. 2. FESEM images of the surface of the lms faced towards the tray during the drying: (a) pristine alginate lm, (b) MFC50, (c) BIRCH50, (d) NFC10, (e) DCC10, (f)
ICBIRCH50 and images of the surfaces of the biocomposite lms faced towards the air during the drying: (g) ICBIRCH50 and (h) ICMFC50.

only a broad halo which indicates amorphous nature (Huq et al.,

2012).
The characteristic cellulose peaks are undetectable in the diffractograms of NFC10 and DCC10 due to the large amorphous peak
of alginate, but they are easily observed in the diffractograms of
MFC50 and BIRCH50. The CrI of cellulose in BIRCH50 is observed
to decrease to 48.3%, making it 17% less crystalline than the pure
birch pulp lm. The CrI of MFC50 is decreased even more, and it
was observed to be 47.2%, making it 34% less crystalline than the
pure MFC lm. This is an unusual phenomenon; however, it has
been proposed that the hydrogen bonding interaction between cellulose crystals and additives can lead to decrease of the crystallinity of cellulose (Ul-Islam, Khan, & Park, 2012). Additives, such as

(a)

DCC10

chitosan, can penetrate cellulose bres and disrupt the hydrogen

bonding of cellulose leading to decrease in crystallinity (Ul-Islam,
Shah, Ha, & Park, 2011). In this context, alginate may penetrate cellulose bres during the lm preparation and interact with crystalline parts of the cellulose, which leads to the decrease of the
crystalline order of cellulose. Furthermore, it is possible the larger
surface area of MFC bres than birch pulp bres allows more efcient penetration of alginate into MFC than birch pulp. This is
shown in the higher decrease of the crystallinity of cellulose in
MFC50 than in BIRCH50.
After ionic cross-linking, there is no change in the X-ray
diffractograms of the pristine alginate lms, NFC10 and DCC10.
However, the crystallinity of the lms containing birch pulp and

(b)

ICDCC10

DCC
ICNFC10
NFC
BRICH50
BIRCH

ICBRICH50

Intensity

Intensity

NFC10

MFC50

ICMFC50

MFC
ICAlg
Alg
20

40

2 (degree)

20

40

2 (degree)

Fig. 3. XRD patterns for the biocomposite lms (a) before and (b) after the cross-linking. XRD patterns pristine alginate and pure birch pulp (BIRCH), MFC, NFC and DCC lms
are also shown.

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

3.5.1. Specic tensile strength and tensile strength

Thicknesses of lms are presented in Supplementary data
(Fig. S2). The thicknesses of the lms increased with the amount
of added cellulose due to the brous structure of cellulose. The difference between the thicknesses of the lms containing the nanoand microbrillated and unbrillated cellulose was enormous, e.g.
the BIRCH50 was 88% thicker than MFC50. The addition of a small
amount of NFC to alginate does not have a signicant impact on
the thickness of lms; however, NFC15 has a 23% higher thickness
compared to the pristine alginate lm. Thicknesses of lms containing DCC were slightly higher compared to the pristine alginate
lms and the lms containing NFC. This suggests that some DCC
agglomerates are formed during the preparation of lms containing DCC and alginate.
After the cross-linking, the thickness of the alginate lms with
birch pulp and MFC is decreased whilst thickness of NFC10 is increased 17% during ionic cross-linking. On the other hand, there
is no signicant change in the thickness of the DCC10 after crosslinking.
An increase of the thickness in the alginate lms during the
cross-linking has been reported previously (Rhim, 2004). The increase is due to the changes in the morphology of the alginate during the cross-linking in the aqueous calcium chloride solution
(Russo, Malinconico, & Santagate, 2007). However, in the case of
ICMFC50 and ICBIRCH50, the decreases in the thicknesses are
probably due to the shrinkage of the alginate which in turn compresses the voids in the biocomposites and makes these lms
denser.
Tensile strengths of the lms are presented in Fig. 4a. To take
into account high differences in the densities of lms, r/q values
have also been reported. The pristine alginate lm had an r/q value of 49.0 kN m/kg and lms with 1% to 15% of MFC have statistically the same r/q values, whilst MFC20 has a 17% and MFC50 a
38% greater r/q values than the pristine alginate lm. On the other
hand, BIRCH50 had practically the same r/q value as the pristine
alginate lm.
Compared to MFC, nano-sized NFC has a clearly better reinforcement effect on the alginate lm. An r/q value of
80.10 kN m/kg was obtained when 10% of NFC was used, which
is 63% and 19% greater than those of MFC10 and MFC50. However,
the increase of NFC content from 10% to 15% did not further enhance r/q of the lms.
Interestingly, DCC1 and DCC10 have practically the same r/q
values as pristine alginate lm. This could be explained by the electrostatic repulsion between the anionic carboxyl groups of alginate
and DCC, which prevents effective interaction between alginate
and DCC, previously observed in FTIR study.
Values of r showed slightly different pattern compared to r/q
values due to the large variations on the densities of lms.
Fig. 4a shows that when the amount of MFC in the composite
was increased, r values of the lms decreased compared to pristine
alginate lms, and MFC10 has a 29% lower r value than pristine
alginate lm. The r value of BRICH50 was also signicantly lower
than the r values of the other lms. As previously noted, based on
the FESEM images, surface of BIRCH50 was uneven, and the pores
on the surface may serve as the breaking points in the lm during
tension.

Specific strength (kNm/kg) and

strength (MPa)

3.5. Mechanical properties

(a) 130
110
90
70
50

kNm
/kg

30

MPa

10

(b)
Specific modulus (MNm/kg) and
modulus (GPa)

MFC are increased, ICMFC50 and ICBIRCH50 having CrI of 55.2%

and 54.2%, respectively. This phenomenon has previously been observed with the blend cotton linter cellulose and alginate membrane after ionic cross-linking with Ca2+ (Yang, Zhang, Peng, &
Zhong, 2000).

20
18
16
14
12
MNm
/kg

10
8

GPa

6
4

(c)

3.5
3.25
3
2.75

Strains (%)

348

2.5
2.25
2
1.75
1.5
1.25
1

Fig. 4. (a) Values of r/q (kN m/kg) and r (MPa) b) E/q (MNm/kg) and E (GPa) and
(c) strains (%) of lms (error bar represents standard deviations).

On the other hand, NFC1 has a 15% higher r value compared to

the pristine alginate lm and by increasing NFC content in the lm
to 15%, the r value is increased 22%. NFC15 had an r value of
97.97 MPa, which is 40% greater that of the pristine alginate lm.
Compared to the MFC lms, NFC lms have signicantly higher r
values, e.g. NFC10 had 63% greater r value than MFC10. It has previously been reported that 8% nanosized cellulose, i.e. nanocrystalline cellulose (NCC) increases the r of alginate lm from 57 to
75.24 MPa (Huq et al., 2012). This and our present studies suggest
that NCC and NFC have quite similar reinforcement effect on alginate lms.
The r value of DCC1 was 13% greater than that of the pristine
alginate lm. However, when the amount of DCC is increased from
1% to 10%, the r value of the lm is decreased from 79.37 to
69.63 MPa. The observation further indicates poor interaction between alginate and DCC.
Mechanical properties of the biocomposite lms were also studied after ionic cross-linking. The slides of ICNFC10 were wrinkled
and uneven which affected on their properties as minor fractures
were formed during the preparation of the slides, and the accurate
results of ICNFC10 was not obtained. On the other hand, ICMFC50,
BIRCH50 and ICDCC10 were smooth and easily processed and they
showed higher r/q values compared to their non-cross-linked

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

counterparts. Highest improvement in r/q was observed with

ICDCC10 which had 54% greater r/q value than DCC10. The r/q value of MFC50 and BIRCH50 increased 13% and 10% after ionic crosslinking, respectively.
The r values of lms were also remarkably increased during
cross-linking and ICMFC50 had a 49% greater r value than
MFC50, and ICDCC10 has a r value of 125.31 MPa which is 80%
greater r compared to DCC10. To the best of our knowledge this
is the highest r of alginate biocomposite lms reported so far
and comparable with cellophane, a commonly used cellulosebased packaging material as r value of cellophane is reported to
be as high as 120 MPa (Spence, Venditti, Habibi, Rojas, & Pawlak,
2010). As previously noted, based on the FTIR studies, the anionic
groups can take part in the formation of cross-linked network during the ionic cross-linking, which results in enhancement of the
mechanical properties of the lm.
Interestingly, the r value of BIRCH50 increased 93% during the
ionically cross-linking. As previously noted, based on the FESEM
images, cross-linking made lm containing birch pulp more even
and reduced the amount of pores on the surface, which reduced
the amount of breaking points in the lm. However, r value of
ICBIRCH50 is still 67% lower that of pristine alginate lm and three
fold lower than the r value of ICDCC10.

349

alginate lm, respectively. Statistically, other lms showed similar

strain values than the pristine alginate lm.
The strain values of the biocomposite lms were slightly decreased during the cross-linking, and strains of the ICMFC50,
ICBIRCH50 and ICDCC10 were 28%, 42% and 7% lower compared
to their non-cross-linked counterparts.
Results from mechanical studies showed that addition of cellulose, especially NFC to the alginate lms increased their strengths
and moduli values, but decreased strain and exibility of the lm.
The improvements in the mechanical properties of composites
containing cellulose bres are due to the formation of tightly
hydrogen bonded networks between bres and matrix.
It was also observed that the chemical nature of the cellulose bre plays important part in the formation of strong biocomposites.
Before cross-linking, the anionic DCC did not have reinforcement
effect on the alginate, due to the electrostatic repulsion between bres and alginate. Even though there are anionic groups in NFC, the
charge density is signicantly lower that of DCC and consequently
NFC had good compatibility with alginate. On the other hand, anionic groups of the DCC were observed to take part in the ionic
cross-linking network after Ca2+ cross-linking, resulting in remarkable increase in the mechanical properties of the lm.
3.6. Barrier properties

3.5.2. Specic modulus and modulus

The E/q values of the lms are presented in Fig. 4b. The pristine
alginate lm had an E/q value of 9.38 M Nm/kg. An increase of the
MFC amount from 1% to 15% does not increase the E/q value of the
biocomposite lm; however, MFC20 and MFC50 had 25% and 47%
greater E/q value than the pristine alginate lm, respectively.
Interestingly, BIRCH50 showed 19% greater E/q value that of the
pristine alginate lm. However, MFC50 had a 24% greater E/q value
than BIRCH50.
The E/q values of the lms containing NFC are consistent with
the r/q values. When amount of NFC was increased, the E/q value
of the biocomposite lm was also increased, and NFC10 had 55%
and 31% higher E/q value than the pristine alginate lm and
MFC10, respectively. No signicant increase in the E/q values can
be observed when amount of NFC is increased from 10% to 15%.
Furthermore, the addition of DCC to alginate lm does not increase
the E/q of lm.
E values of the lms shows slight decrease when the amount of
MFC was increased. However, statistically all of the MFC lms
showed almost same E that of the pristine alginate lm. Similarly
to the r values, the E value of BIRCH50 is also signicantly lower
than the E values of the other lms.
Addition of the NFC to the alginate lm increased E values of the
lm and NFC1 had 13% and NFC5 27% higher E values than the pristine alginate lm. In addition, NFC10 had 51% higher E value than
MFC10. Statistically there was no difference in E values between
the pristine alginate lm and lms containing DCC bres.
The E/q value of MFC50 was not affected by ionic crosslinking;
however, the E value of ICMFC50 was 30% lower than the E value of
MFC50. The E/q value of ICBIRCH50 was also almost same as the
E/q value of BIRCH50 and the E value of BIRCH50 was actually
increased 69% after cross-linking. In the biocomposite lms with
10% of DCC, the E/q and the E values were increased 21% and
41%, respectively, after ionic cross-linking.
3.5.3. Elongation at break
Values of the strain of the lms are presented in Fig. 4c. It is
known that incorporation of nanocellulose bres to other polysaccharide matrixs decreases strain of the lms (Fernandes et al.,
2010). Films with 10% and 15% NFC showed signicant decrease
in strain and they had 66% and 61% lower strains than the pristine

3.6.1. Grease barrier properties

All of the studied cellulose-containing alginate lms except
BIRCH50 had excellent grease barrier properties, i.e. the dyed turpentine did not penetrate lms within the 30 min testing period.
However, the dyed turpentine penetrated BIRCH50 immediately
which was probably due to porous structure of the lm previously
noted, based on the FESEM images. On the other hand, surfaces of
the other lms were smooth and visible pores were not observed
based on FESEM images. Moreover, the good grease barrier properties of alginate explain partly the excellent grease barrier properties of the biocomposite lm (Ham-Pichavant, Sbe, Pardon, &
Coma, 2005).
After ionic cross-linking, lms retained their grease barrier
properties and no penetration of the dyed turpentine was observed
in ICNFC10, ICDCC10 and in ICMFC50. However, penetration of the
dyed turpentine occurred immediately with ICBIRCH50.
3.6.2. Water vapour permeability
The WVP of the pristine alginate lm, BIRCH50, MFC50, NFC10
and DCC10 was measured before and after ionic crosslinking
(Fig. 5a). The WVP of pristine alginate lm was 8.4  107 g/
(Pa s m2). Due to the high hydrophilicity, alginate has low water
vapour barrier properties (Olivas & Barbosa-Cnovas, 2008). The
addition of micro- and nanocellulose bres to the alginate lm decreased the WVP, e.g. the WVP of the MFC50 was a 23% lower than
that of the pristine alginate lm. BIRCH50 had in turn a 6% higher
WVP that of the pristine alginate lm.
It is known that barrier properties of lms are improved when
additives have lower permeability and materials are well dispersed
(Khan et al., 2010). Furthermore, addition of micro- and nano-sized
cellulose bres to the alginate matrix makes alginate lms more
tortuous, and thus decreases the pathways to water molecules to
pass through the lm (Huq et al., 2012). This could explain the improved water vapour barrier properties of alginate lms after addition of MFC, DCC and NFC bres. The large pores, observed in
FESEM images, may create direct pathways for water vapour to
penetrate through BIRCH50, resulting in decreased water vapour
barrier properties compared to the pristine alginate lm.
It is known that ionic cross-linking improves water vapour barrier properties of alginate and composites containing alginate
(Rhim, 2004). Here water vapour barrier properties were also

350

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

a cross-linking. On the other hand, anionic carboxylate groups

created a highly intense cross-linked network between alginate
and DCC, when the lm was ionically cross-linked with Ca2+. A biocomposite lm with remarkably high mechanical strength was
produced by ionically cross-linking the alginate lm containing
10% of DCC. All of the biocomposite lms except the lm with
unmodied birch pulp, showed excellent grease barrier properties.
Furthermore, the WVP of alginate lm was decreased by the addition of MFC and nano-sized cellulose bres, and by ionic cross-linking. This work demonstrates that biocomposite lms based on
cellulose bres and alginate have high potential to be used as high
strength packaging materials.
Acknowledgments
The authors acknowledge the support from the Future
Biorenery Programme of Forestcluster Ltd. V.T.T. is acknowledged
for the pulp analysis.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foodchem.2013.
11.037.
Fig. 5. (a) WVP (g/(Pa s m2)) and (b) SWVP (kg m/(Pa s m2)) of lms (error bar
represents standard deviations).

observed to improve after ionic cross-linking and both ICNFC10

and ICDDC10 have a 13% lower WVP than the pristine alginate lm.
Furthermore, ICNFC10 has a 6% lower WVP than NFC10 and
ICDD10 has a 9% lower WVP than DCC10.
WVP of MFC50 was increased after cross-linking, and ICMFC50
has a 14% higher WVP than MFC50. However, the WVP of ICMFC50
is still a 6% lower than the pristine alginate lm. On the other hand,
the water vapour barrier properties of BIRCH50 was decreased
after cross-linking and ICBIRCH50 has 11% and 5% lower WVP compared to BIRCH50 and the pristine alginate lm, respectively. As
previously noted, based on the FESEM images, the number of pores
in the surface of BIRCH50 was reduced during the cross-linking.
This in turn reduced the direct pathways of water vapour to penetrate through the lm.
To take into account the differences in the thicknesses of lms,
SWVPs are presented in Fig. 5b. There were no signicant differences in the SWVPs of the lms, expect in the lms containing
birch pulp, i.e. SWVP of BRICH50 is more than threefold
higher than that of the pristine alginate lm. After cross-linking,
SWVP of the BIRCH50 decreased 32%, but was still 91% higher
compared to pristine alginate lms. Compared to common
packaging materials, such as cellophane, all of the biocomposite
lms had lower SWVP than cellophane whose SWVP is reported
to be 6.9  1014 kg m/(Pa s m2) (Proval, Debeaufort, Despre, &
Voilley, 2002).
4. Conclusion
In this work, biocomposite lms based on cellulose and alginate
were produced using four different wood cellulose bres with
different bre sizes, chemical nature and proportions. Films were
also further cross-linked using Ca2+. Celluloses showed good interaction with alginate matrix and the mechanical properties of lms
were increased by increasing the amount of the bre in the
biocomposite and by decreasing the bre size of cellulose. However, despite the nanosized bre structure, anionic charges prevented an efcient interaction between DCC and alginate without

References
Abdul Khalil, H. P. S., Bhat, A. H., & Ireana Yusra, A. F. (2012). Green composites from
sustainable cellulose nanobrils: A review. Carbohydrate Polymers, 87, 963979.
Agulhon, P., Markova, V., Robitzer, M., Quinard, F., & Mineva, T. (2012). Structure of
alginate gels: Interaction of diuronate units with divalent cations from density
functional calculations. Biomacromolecules, 13, 18991907.
ASTM Standard E 9695 (1995). Standard test for water vapor transmission of
materials, Philadelphia, USA.
Ayranci, E., & Tunc, S. (2001). The effect of fatty acid content on water vapour and
carbon dioxide transmissions of cellulose-based edible. Food Chemistry, 72,
231236.
Benavides, S., Villalobos-Carvajal, R., & Reyes, J. E. (2012). Physical, mechanical and
antibacterial properties of alginate lm: Effect of the crosslinking degree and
oregano essential oil concentration. Journal of Food Engineering, 110, 232239.
Bhattacharya, A., Rawlins, J. W., & Ray, P. (2009). Polymer grafting and crosslinking.
New Jersey: John Wiley & Sons, Inc., p. 3.
Cheng, L. H., Abd Karim, A., & Seow, C. C. (2008). Characterisation of composite lms
made of konjac glucomannan (KGM), carboxymethyl cellulose (CMC) and lipid.
Food Chemistry, 107, 411418.
Cheng, Y., Lu, L., Zhang, W., Shi, J., & Cao, Y. (2012). Reinforced low density alginatebased aerogels: Preparation, hydrophobic modication and characterization.
Carbohydrate Polymers, 88, 10931099.
Fernandes, S. C. M., Freire, C. S. R., Silvestre, A. J. D., Neto, C. P., Gandini, A., Berglund,
L. A., et al. (2010). Transparent chitosan lms reinforced with a high content of
nanobrillated cellulose. Carbohydrate Polymers, 81, 394401.
Fuks, L., Filipiuk, D., & Majdan, M. (2006). Transition metal complexes with alginate
biosorbent. Journal of Molecular Structure, 792793, 104109.
Ham-Pichavant, F., Sbe, G., Pardon, P., & Coma, V. (2005). Fat resistance properties
of chitosan-based paper packaging for food applications. Carbohydrate Polymers,
61, 259265.
Hansen, N. M. L., & Plackett, D. (2008). Sustainable lms and coatings from
hemicelluloses: A review. Biomacromolecules, 9, 14931505.
Huq, T., Salmieri, S., Khan, A., Khan, R. A., Tien, C. L., Riedl, B., et al. (2012).
Nanocrystalline cellulose (NCC) reinforced alginate based biodegradable
nanocomposite lm. Carbohydrate Polymers, 90, 17571763.
ISO Standard 5273 (1995). Plastics determination of tensile properties Part 3:
Test conditions for lms and sheets, Geneve, Switzerland: ISO.
Iwamoto, S., Nakagaito, A. N., & Yano, H. (2007). Nano-brillation of pulp bers
for the processing of transparent nanocomposites. Applied Physics A, 89,
461466.
Johnson, F. A., Craig, D. Q. M., & Mercer, A. D. (1997). Characterization of the block
structure and molecular weight of sodium alginates. Journal of Pharmacy and
Pharmacology, 49, 639643.
Johnson, R. K., Zink-Sharp, A., Renneckar, S. H., & Glasser, W. G. (2009). A new biobased
nanocomposite:
Fibrillated
TEMPO-oxidized
celluloses
in
hydroxypropylcellulose matrix. Cellulose, 16, 227238.
Khan, R. A., Salmieri, S., Dussault, D., Uribe-Calderon, J., Kamal, M. R., Safrany, A.,
et al. (2010). Production and properties of nanocellulose-reinforced
methylcellulose-based biodegradable lms. Journal of Agriculture and Food
Chemistry, 58, 78787885.

J.A. Sirvi et al. / Food Chemistry 151 (2014) 343351

Kristo, E., Biliaderis, C. G., & Zampraka, A. (2007). Water vapour barrier and tensile
properties of composite caseinate-pullulan lms: Biopolymer composition
effects and impact of beeswax lamination. Food Chemistry, 101, 753764.
Liimatainen, H., Sirvi, J., Haapala, A., Hormi, O. E., & Niinimki, J. (2011).
Characterization of highly accessible cellulose microbers generated by wet
stirred media milling. Carbohydrate Polymers, 83, 20052010.
Liimatainen, H., Visanko, M., Sirvio, J. A., Hormi, O. E., & Niinimaki, J. (2012).
Enhancement of the nanobrillation of wood cellulose through sequential
periodatechlorite oxidation. Biomacromolecules, 13, 15921597.
Lin, N., Bruzzese, C., & Dufresne, A. (2012). TEMPO-oxidized nanocellulose
participating as crosslinking aid for alginate-based sponges. ACS Applied
Material Interfaces, 4, 49484959.
Mourio, V., Newby, P., Pishbin, F., Cattalini, J. P., Lucangioli, S., & Boccaccini, A. R.
(2011). Physicochemical, biological and drug-release properties of gallium
crosslinked alginate/nanoparticulate bioactive glass composite lms. Soft
Matter, 7, 67056712.
Nakamura, K., Nishimura, Y., Hatakeyama, T., & Hatakeyama, H. (1995). Thermal
properties of water insoluble alginate lms containing di- and trivalent cations.
Thermochimica Acta, 267, 343353.
Olivas, G. I., & Barbosa-Cnovas, G. V. (2008). Alginatecalcium lms: Water vapor
permeability and mechanical properties as affected by plasticizer and relative
humidity. LWTFood Science and Technology, 41, 359366.
Papageorgiou, S. K., Kouvelos, E. P., Favvas, E. P., Sapalidis, A. A., Romanos, G. E., &
Katsaros, F. K. (2010). Metal-carboxylate interactions in metal-alginate
complexes studied with FTIR spectroscopy. Carbohydrate Research, 345,
469473.
Proval, C., Debeaufort, F., Despre, D., & Voilley, A. (2002). Edible arabinoxylanbased lms. 1. Effects of lipid type on water vapor permeability, lm structure,
and other physical characteristics. Journal of Agriculture and Food Chemistry, 50,
39773983.
Ramesh, A., Hasegawa, H., Sugimoto, W., Maki, T., & Ueda, K. (2008). Adsorption of
gold(III), platinum(IV) and palladium(II) onto glycine modied crosslinked
chitosan resin. Bioresource Technology, 99, 38013809.
Rhim, J. W. (2004). Physical and mechanical properties of water resistant sodium
alginate lms. LWTFood Science and Technology, 37, 323330.

351

Russo, R., Malinconico, M., & Santagate, G. (2007). Effect of cross-linking with
calcium ions on the physical properties of alginate lms. Biomacromolecules, 8,
31933197.
Segal, L., Creely, J. J., Martin, A. E., Jr, & Conrad, C. M. (1959). An empirical method for
estimating the degree of crystallinity of native cellulose using the x-ray
diffractometer. Textile Research Journal, 29, 786794.
Sikareepaisan, P., Ruktanonchai, U., & Supaphol, P. (2011). Preparation and
characterization of asiaticoside-loaded alginate lms and their potential for
use as effectual wound dressings. Carbohydrate Polymers, 83, 14571469.
Sikorski, P., Mo, F., Skjk-Brk, G., & Stokke, B. T. (2007). Evidence for egg-boxcompatible interactions in calcium-alginate gels from ber X-ray diffraction.
Biomacromolecules, 8, 20982103.
Spence, K. L., Venditti, R. A., Habibi, Y., Rojas, O. J., & Pawlak, J. J. (2010). The effect of
chemical composition on microbrillar cellulose lms from wood pulps:
Mechanical processing and physical properties. Bioresource Technology, 101,
59615968.
Sui, K., Li, Y., Liu, R., Zhang, Y., Zhao, X., Liang, H., et al. (2012). Biocomposite ber of
calcium alginate/multi-walled carbon nanotubes with enhanced adsorption
properties for ionic dyes. Carbohydrate Polymers, 90, 399406.
TAPPI Standard T 454 om-10 (2010). Turpentine test for voids in glassine and
greaseproof papers, TAPPI Press, Atlanta.
Ul-Islam, M., Khan, T., & Park, J. K. (2012). Nanoreinforced bacterial cellulosemontmorillonite composites for biomedical applications. Carbohydrate
Polymers, 89, 11891197.
Ul-Islam, M., Shah, N., Ha, J. H., & Park, J. K. (2011). Effect of chitosan penetration on
physico-chemical and mechanical properties of bacterial cellulose. Korean
Journal of Chemical Engineering, 28, 17361743.
Yang, G., Zhang, L., Peng, T., & Zhong, W. (2000). Effects of calcium(II) bridge
crosslinking on structure and pervaporation of cellulose/alginate blend
membranes. Journal of Membrane Science, 175, 5356.
Zoppe, J. O., Peresin, M. S., Habibi, Y., Venditti, R. A., & Rojas, O. J. (2009). Reinforcing
poly(e-caprolactone) nanobres with cellulose nanocrystals. ACS Applied
Material Interfaces, 1, 19962004.