Materials Letters 64 (2010) 901904

Contents lists available at ScienceDirect

Materials Letters
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / m a t l e t

Bacterial cellulosepoly(vinyl alcohol) nanocomposites prepared by an

in-situ process
S. Gea a,b, E. Bilotti a,b, C.T. Reynolds a,b, N. Soykeabkeaw a,b,c, T. Peijs a,b,d,
a
Centre for Materials Research, Queen Mary University of London, London E1 4NS, UK
b
School of Engineering and Material Science, Queen Mary University of London, London E1 4NS, UK
c
Mae Fah Luang University, 333 Moo 1, School of Science, Thasud Muang, Chiang Rai 57100, Thailand
d
Eindhoven University of Technology Eindhoven Polymer Laboratories, PO Box 513, 5600 MB Eindhoven, The Netherlands

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

Article history: Bio-nanocomposites were prepared by an in-situ growth process through the direct addition of poly(vinyl
Received 2 November 2009 alcohol) (PVA) into the Acetobacter xylinum inoculated medium and compared with composites made by
Accepted 13 January 2010 impregnation of bacterial cellulose (BC) gels with a PVA solution. Mechanical property tests showed that the
Available online 21 January 2010
presence of PVA in the BC acts as a plasticizer, interrupting hydrogen bonding between cellulose brils
within the BC network. This resulted in a reduction in Young's modulus and an increase in toughness
Keywords:
compared to pure BC sheet, especially for in-situ grown samples. Interestingly, the small amount of added
Poly(vinyl alcohol)
Bacterial cellulose nanocomposites
PVA turns the BC sheet into optical transparent nanocomposite lms with excellent mechanical properties.
In-situ growth 2010 Elsevier B.V. All rights reserved.
Impregnation
Mechanical properties
Transparent lm

1. Introduction signicant improvement in mechanical properties [4]. High volume

Research into cellulose synthesised by micro-organisms has
received a great deal of attention in recent decades [1]. Some bacteria
such as Acetobacter xylinum are able to synthesise nanobrils of
bacterial cellulose (BC) through the polymerisation of glucose
molecules converted into -1,4-glucan chains in the interior of BC
[2]. In static culture conditions, these bacteria produce a thick gel or
pellicle which has unique properties such as high purity, high
crystallinity, remarkable mechanical properties and an ability to
form homogenous membrane sheets [2,3,4]. As this nanosize mat
possesses excellent biocompatibility and an ultrane reticulated
structure, BC has found a multitude of applications in the paper,
textile, and food industries, and as a biomaterial in cosmetics and
medicine [5].
Various efforts have been made to maximise the benets of
nanocomposites based on BC, either by combining the synthesized
pellicles with other materials or by modifying the cellulose biosyn-
thesis [1]. Nanocrystals of BC, which are obtained by hydrolysing the
amorphous part of BC, have been applied as a reinforcing agent in
cellulose butyrate acetate [6], and starch [7]. By disintegrating the BC
gel in the wet state, in order to easily blend it with other cellulose
bres, paper prepared from apple and radish pulp showed a
Corresponding author. Centre for Materials Research, Queen Mary University of
London, London E1 4NS, UK. Tel.: + 44 20 878828865.
E-mail address: t.peijs@qmul.ac.uk (T. Peijs).
fraction composites were prepared by Yano et al. by impregnating BC
sheets with thermoset resins [8]. These materials showed a high
stiffness in combination with optical transparency, but were quite
brittle. For applications where ductility and toughness are required,
ductile polymer matrices are more of interest. A membrane for the
pervaporative separation of ethanol/water has been prepared by
impregnating BC with 1.5% chitosan solution in 1% aqueous acetic acid
[9]. More recently, an all-cellulose nanocomposites entirely made
from BC was successfully prepared by immersing a BC sheet in solvent
(lithium chloride/N,N-dimethylacetamide [10].
Alternatively, various substances have been directly incorporated
into the cultural medium, via an in-situ growth process. For use in the
healing of wounds, ulcers and burns, chitosan was added to the
culture medium [11]. Carboxymethylated-bacterial cellulose, as
copper and lead ion removal membrane, has been obtained by adding
the water-soluble carboxymethylated cellulose directly into the
culture medium [12]. BC was also grown in the presence of potato
and corn starch [13], and poly(vinyl alcohol) (PVA). PVA, in particular,
has attracted a lot of interests, because of its water solubility, good
mechanical properties and biocompatibility [14]. It was found that the
presence of PVA can inuence the in-situ growth of BC through the
formation of spherulites [15]. Seifert et al. found that the addition of
cellulose derivatives and PVA into the culture medium gave
biocomposites with better water retention and ion absorption [16].
In this study, two different preparation methods for BC/PVA
nanocomposites are compared; (i) in-situ growth of BC in the

0167-577X/$ see front matter 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.matlet.2010.01.042
902 S. Gea et al. / Materials Letters 64 (2010) 901904

presence of PVA (In-situ BC) and (ii) impregnation of bacterial

cellulose gels with a PVA solution (Impreg BC). The two composites
obtained are characterised for what concerns their morphological and
mechanical properties.

2. Experimental

2.1. Materials and methods

Bacterial cellulose was prepared according to a method described

previously [2,4]. A concentration of 5% v/v PVA (Mw = 84,000
124,000 g/mol; 9899% hydrolysed) (Aldrich, USA) solution in
distilled water was prepared by stirring for 30 min at 80 C. A sample
of 100 g puried BC gel was impregnated with 100 ml of the PVA
solution overnight and then pressed (115 C for 5 min at 70 MPa) into
a sheet. This sample is hereafter called Impreg BC. The same
concentration of (sterilised) PVA solution (100 ml) was alternatively
added directly into the medium (medium B in [4]) of the growing BC.
This sample is referred to as In-situ BC. The In-situ BC was washed in
fresh water and then left overnight on a porous wire sieve to remove
the water. As a control, sheets of pure bacterial cellulose (simply
referred to as BC) gel and PVA were also prepared.
It should be noted that, although the initial PVA concentration for
the preparation of In-Situ BC and Impreg BC is similar, the nal
amount of PVA present in the two composites is different, because of
the different purication steps. The estimated PVA matrix content in
Impreg BC composites is 3.7%, while the PVA content in the In-situ BC
composites is 1.4% before purication.

2.2. Characterization

The morphology of BC/PVA composites was observed using a JEOL

JSM-6300F Field Emission Scanning Electron Microscope (SEM). All
specimens were coated with gold before SEM observation.
Tensile tests were performed on a tensile testing machine (Instron
5564) at a cross-head speed of 5 mm/min using a 1 kN static load cell.
The tensile test specimens were rectangular strips (40 mm 5 mm),
cut from 0.3 mm thick sheets, and were tested at ambient conditions,
after been dried overnight at 80 C in vacuum. All measurements were
performed for at least ve samples in each case and the average value
was recorded.
Structural characterisation was carried out using a Siemens D5000
X-ray diffractomer (40 kV, 40 mA) with Cu K radiation ( = 0.154 nm)
was used. The aperture slits were set at 0.1 and the scanning step was
0.02 with a scan time of 2.5 s per step.

3. Results and discussion

Since any chemicals added to the medium can affect the BC growth
[1620], the morphology of the in-situ grown BC was studied. Fig.1
shows the morphology of BC and BC/PVA bio-nanocomposites
prepared by the different methods.
The morphology of BC nanobres (Fig.1a) is characterized by a
network of interconnected brils that contain entanglements and
aggregations [4]. The preparation of BC by an in-situ process with a
solution of PVA 5%v/v directly added to the culture medium did not
signicantly affect the morphology of the bacterial cellulose. Apart
from the fact that the diameter of the BC brils can be seen to change
slightly, no signicant changes occur in the arrangement of the brils.
Impregnation of PVA 5%v/v into a BC gel has however some effect on
the ribbon morphology of BC. Aggregations form more easily and
thicken in certain parts, resulting in an increase in the diameter of the
BC bre bundles. The same can be seen also for Impreg BC, where the
bres bundles are even more slightly thicker than those of In-situ BC.
From the SEM images of Fig. 1(b) and (c) it can also be seen that the
Fig. 1. SEM micrographs of: (a) BC (Magnication 5000), (b) In-situ BC (Magnication
8000) and (c) Impreg BC composites (Magnication 5000), showing the presence of
PVA on the BC ribbons and the more homogeneous network for In-situ BC.
amount of PVA in Impreg BC is higher than in In-situ BC and less
homogeneously distributed.
The different morphologies found are expected to signicantly
inuence the mechanical performances. In general, the presence of an
external phase between cellulose bres tends to weaken the strong
attractive hydrogen bonding between them [3,10], affecting the
mechanical properties of the cellulose lms. In the case of composites
made by the impregnation method the presence of PVA reduces both
the tensile strength and stiffness of the lms by a factor of two, with
the Young's modulus decreasing from 18.8 GPa (BC) to 9.1 GPa
(Fig. 2).
The in-situ grown BC/PVA composite, instead, provides a better
combination of the characteristics of BC and PVA, notably high
strength and ductility. At rst the composite responds elastically,
 S. Gea et al. / Materials Letters 64 (2010) 901904 903

Fig. 2. Stressstrain curves of pure BC compared with In-situ BC and Impreg BC composites. The inset graph shows in greater detail the stressstrain behaviour at small deformations.

with a modulus (6.4 GPa) closer to the one of pure PVA (3.0 GPa) than
that of BC (18.8 GPa). This is followed by a plastic deformation process
with necking and, successively, a strain hardening regime with a
modulus of 14.7 GPa, close to that of pure BC. Moreover, the ultimate
tensile stress of the in-situ BC/PVA composite is similar to the value of
pure BC, suggesting that the presence of the small, and homogeneously
dispersed, amount of PVA does not compromise the ultimate strength
of the BC network. Importantly, the strain at break for the in-situ BC/
PVA composite is signicantly higher than that of pure BC sheet,
providing the necessarily ductility to these materials. The presence of a
homogeneous PVA coating onto the individual BC nanobres leads to
nanobre slippage within the BC network due to the partial
interruption of hydrogen bonds between BC, allowing the composites
to feature an attractive combination of strength, modulus and work-
of-fracture. Initially, the material acts as a composite of PVA coated BC
while at high deformations the plastically deformed network of BC
takes over.
Fig. 3 shows that the diffractograms of BC, In-situ BC and Impreg
BC are practically identical. The three peaks located at 2 = 14.7,
16.2, and 22.4 correspond to the primary diffraction of the (110),
(110), and (200) planes of polymorph cellulose I [21]. These data are
characteristic of BC and comparable with X-ray diffractograms
reported previously [3,10,22]. From these data it can be concluded
that the presence of PVA has no effect on the cellulose crystal
structure of either In-situ BC or Impreg BC. However, a difference can
be seen in the relative intensity of the peaks, especially in In-situ BC
and Impreg BC, suggesting a change in the orientation of the cellulose
bres in the presence of PVA.
Fig. 4 shows the effect of the presence of PVA in the BC lm on the
transparency of the two composites: In-situ BC and Impreg BC. The

Fig. 3. X-ray diffractogram of pure BC, In-situ BC and Impreg BC, showing no effect of the presence of PVA on the crystal structure of In-situ BC and Impreg BC.
904 S. Gea et al. / Materials Letters 64 (2010) 901904
Fig. 4. Optical photograph of (a) pure BC, (b) In-situ BC, (c) Impreg BC, and (d) pure PVA
sheet, showing transparency after the introduction of PVA.
addition of PVA into BC leads to an improvement in the optical
transparency of the pressed lm. The more PVA trapped in the BC the
greater the optical transparency. As a result, the optical transparency
of Impreg BC, with an estimated PVA content of 3.7 %, is better than
that of In-situ BC, with 1.3 % maximum PVA content. This
phenomenon is due to the effect of the nanosize bres of BC used in
this work, which have the ability to retain the transparency of PVA
[23]. The empty space in BC lm is easily lled by PVA [8] and has a
good wetting and interfacial bonding with BC and similar refractive
index.
4. Conclusions
It has been shown that 5% v/v water-soluble PVA in aqueous
solution can be added to a culture medium of A. xylinum without
affecting the crystal structure of the resulting BC. The addition of PVA
causes a minor change in the dimensions of BC ribbons, as well as
leading to a reduction in Young's modulus but an increase in ductility
and work-of-fracture. Especially in the case of in-situ grown
composite lms these materials show high strength, stiffness and
toughness. The increased ductility is due to the partial interruption of
hydrogen bonds which naturally occurs in BC nanobre networks. The
small amount of PVA added to the BC gel, coats the BC bres and leads
to enhanced transparency of the BC/PVA composites. The in-situ
growth technique resulted in a BC/PVA nanocomposite with a better
combination of mechanical and optical properties due to a more
homogeneous microstructure while, at the same time, avoiding a
subsequent impregnation process for the fabrication of the composite.
References
[1] Eichhorn SJ, Dufresne A, Aranguren A, Marcovich NE, Capadona JR, Rowan SJ, et al.
J Mater Sci 2009;45(1):133.
[2] Iguchi M, Yamanaka S, Budhiono. J Mater Sci 2000;35:26170.
[3] Yamanaka S, Watanabe K, Kitamura N. J Mater Sci 1989;24:31415.
[4] Gea S, Torres FG, Tronscoso OP, Reynolds CT, Vilasecca F, Iguchi M, Peijs T. Int
Polym Process 2007;22:497501.
[5] Bielecki S, Krystinowicz A, Turkiewcz MBC, Kalinowska H. Bacterial cellulose, in:
Polysaccharides and polyamide in the food industry: Properties, production and
patents, Wiley VCH;2005, Weinhein.
[6] Grunert M, Winter WT. J Polym Environ 2002;10(12):2730.
[7] Orts WJ, Shey J, Imam SH, Glenn GM, Guttman ME, Revol JF. J Polym Environ
2005;13(4):3016.
[8] Yano H, Sugiyama J, Nakagaito AN, Nogi M, Hikita M, Handa K. Adv Mater 2005;17:
1535.
[9] Dubey V, Pandey LK, Saxena C. J Membr Sci 2005;251:1316.
[10] Soykeabkaew N, Sian C, Gea S, Peijs T. Cellulose 2009;16:43544.
[11] Ciechanska D. Fibres Text East Eur 2004;12:6972.
[12] Chen S, Zou Y, Yan Z, Shen W, Shi S, Zhan X, et al. J Hazard Mater 2009;161:13559.
[13] Grande CJ, Torres FG, Gomez CM, Tronscoso OP, Canet Ferrer J, Martinez-Pastor.
J Sci Eng C 2009;29:1098104.
[14] Sun P, Chen J, Liu ZW, Liu ZT. J Macromol Sci, Part A: Pure Appl Chem 2009;46:
53340.
[15] Kai A, Kobayashi T. Polym J 1992;24:1313.
[16] Seifert M, Hesse S, Kabrelian V, Klemm D. J Polym Sci: Part A: Polym Chem
2004;42:46370.
[17] Brown E, Laborie MPG. Biomacromol 2007;8:307481.
[18] Haigler CH, Chanzy HJ. J Ultrastruct Mol Struct Res 1984;98:299311.
[19] Yamamoto H, Horii F, Hirai A. Cellulose 1996;3:22942.
[20] Sugiyama J, Vuong R, Chanzy H. Macromol 1991;24:416875.
[21] Klechkovskaya VV, Baklagina YB, Stepina ND, Buffat PA, Khripunov AK, Suvorova
EI, Tkchenko AA, Zanaveskina IS, Gladchenko SV. Crystal Rep 2003;48:75562.
[22] Zugenmaier P. Crystalline cellulose and derivatives: characterization and
structure. New York: Springer; 2008.
[23] Iwamoto S, Nakagaito AN, Yano H, Nogi M. Appl Phys A 2005;81:110912.