Materials Science and Engineering C 29 (2009) 1098–1104
Contents lists available at ScienceDirect
Materials Science and Engineering C j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m s e c
Development of self-assembled bacterial cellulose–starch nanocomposites Cristian J. Grande a, Fernando G. Torres a,⁎, Clara M. Gomez b,⁎, Omar P. Troncoso a, Josep Canet-Ferrer c, Juan Martínez-Pastor c a b c
Faculty of Mechanical Engineering, Catholic University of Peru (PUCP), Lima 32, Peru Departament de Química Física and Institut de Ciencia dels Materials, Dr Moliner 50, Universitat de València, E-46100 Burjassot, Valencia, Spain Unit of Optoelectronic Materials and Devices of the University of Valencia, P.O. Box 22085, 46071 Valencia, Spain
a r t i c l e
i n f o
Article history: Received 8 February 2008 Received in revised form 5 May 2008 Accepted 11 September 2008 Available online 22 September 2008 Keywords: Self-assembled Bottom-up Bacterial cellulose Starch
a b s t r a c t A bioinspired bottom-up process was developed to produce self-assembled nanocomposites of cellulose synthesized by Acetobacter bacteria and native starch. This process takes advantage of the way some bacteria extrude cellulose nanofibres and of the transport process that occurs during the gelatinization of starch. Potato and corn starch were added into the culture medium and partially gelatinized in order to allow the cellulose nanofibrils to grow in the presence of a starch phase. The bacterial cellulose (BC)–starch gels were hot pressed into sheets that had a BC volume fraction higher than 90%. During this step starch was forced to further penetrate the BC network. The self-assembled BC–starch nanocomposites showed a coherent morphology that was assessed by Atomic Force Microscopy (AFM) and Environmental Scanning Electron Microscopy (ESEM). The nanocomposites structure was studied using X-ray diffraction and ATR-FTIR spectroscopy. The degree of crystallinity of the final nanocomposites was used to estimate the volume fraction of BC. The aim of this paper is to explore a new methodology that could be used to produce nanomaterials by introducing a different phase into a cellulose nanofibre network during its assembly. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Self-assembling materials are designed to organise spontaneously and hierarchically into complex structures in appropriate environments [1,2]. In nature, all materials are formed by using building blocks that aggregate themselves into the desired structures. One example of these self-assembled structures can be found in the cellulose produced by Acetobacter bacteria which is formed by a hierarchical cell-directed self-assembly process. In fact, bacterial cellulose (BC) is extruded through the bacteria cell pores in the form of ribbons. These ribbons start at specific points on the cell surface and become thicker as they build a composite ribbon. Finally, the cellulose nanofibres (2–4 nm in diameter and several 100 μm in length [3–6]) aggregate on the top of the culture medium, incorporate water and form a three dimensional coherent network (BC gel). Such network gels have been processed and used in the past in headphone diaphragms, electroacustic transducers, high strength paper, components for precision optical devices [4,7], etc. Reported biomedical applications of bacterial cellulose include its use as wound dressing [8], guided tissue regeneration [9,10] and scaffolds for tissue engineering [11–15].
⁎ Corresponding authors. E-mail addresses:
[email protected] (F.G. Torres),
[email protected] (C.M. Gomez). 0928-4931/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.msec.2008.09.024
BC differs from plant cellulose in its higher purity, crystallinity (above 60%), degree of polymerization (between 2000 and 6000) and tensile strength [4,5,16,17]. It has been determined that BC has Iα and Iβ crystalline forms unlike the cellulose of plants that mainly presents the Iβ structure [18]. BC exhibits remarkable mechanical properties due to the uniform ultrafine-fibre network structure, the high planar orientation of the ribbon-like fibres when compressed into sheets, the good chemical stability and the high water absorption capacity [4,19,20]. Most BC nanocomposites reported in the literature have been produced by blending the matrix and the second phase after the nanofibres network has been formed. Some authors have tried to disintegrate the cellulose network structure in order to blend it as standard nanofiller [4–7] while other studies have reported the introduction of a second phase without the disintegration of the cellulose network [21,22]. A different approach, much less investigated than the previous one, consists in introducing a second phase during the development of the BC network as in the work of Touzel et al. [23] where a double network gel of pectin-BC has been obtained by adding pectin to the culture medium. This type of approach has been also used by Mormino et al. [24] to incorporate solid reinforcing particles during formation of the BC gel by means of a rotating disc bioreactor. A common feature of these two investigations is that the phase incorporated to the BC network can neither be controlled nor modified. The aim of this paper is not to produce a composite material to improve the mechanical properties of a polymeric matrix, but to develop a technique for the production of self-assembled nanocomposites. This
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
technique uses the natural bottom-up process found in the synthesis of bacterial cellulose in order to produce a self-assembled nanocomposite by aggregating native starch into the nanofibres network. Starch undergoes a process called gelatinization (further described in the next section) which allows to partially controlling the diffusion of polymeric chains into an existing fibre network, in this case, BC. This “smart” property of starch is a key issue in the production of these bioinspired nanocomposites, with a high volume fraction of the strong phase, namely, nanofibres of BC, covered by a starch phase. Considering that both, starch and BC, have potential advantages as biomaterials [8–14,22,25–28], the nanocomposites produced with the method described here, could also have a variety of potential biomedical applications. The technique presented here could be extended for the addition of other materials into the BC network. Further work including the use of this technique in the production of BC–hydroxyapatite nanocomposites for biomedical applications [28] and other systems is being carried out in our laboratories. 2. Bottom-up manufacturing technique The process is based on the addition of intact starch granules to the culture medium before bacteria are inoculated. After being added to the culture medium (Fig. 1a), the starch-medium suspension is autoclaved at 121°. During this process the starch phase, undergoes a “first gelatinization”. The starch granules swell, leach amylose and lose their crystalline structure by the melting of their ordered regions [29]. At this stage bacteria are inoculated and the nanofibres network is formed in the presence of a partially gelatinized starch structure (Fig. 1c). Starch granules are semicrystalline and contain two polymers, i.e., branched amylopectin with a weight-average molecular weight (MW) of 0.7–57×108 and nearly linear amylose (MW 0.3–1.9×106 [30]). The contents of amylose in granules of native starch are in the range 15–30%. The crystalline properties of starch are associated with the short-chain fraction of amylopectin arranged as doubles helices and packed in crystallites [31]. Different carbon sources can be used for the cellulose synthesis, namely glucose, fructose or gluconates [32,33]. Haigler et al. [34] reported that the starch added in the culture medium of Acetobacter xylinum did not affect the formation of cellulose. Thus, it is reasonable to expect that starch is not consumed by Acetobacter bacteria, but instead it remains present in the final nanocomposite. A simple way of revealing the presence of starch amylose is by adding a drop of iodine to a substrate. If it leaves a blue stain, then, it can be inferred that amylose is present [35,36]. This was the case for the BC–starch nanocomposites reported here.
1099
As a final step, in order to produce BC–starch nanocomposite sheets, the BC–starch gels are hot pressed (Fig. 1d) and the starch phase undergoes a “second gelatinization”, which is probably associated to the remaining intact starch granules and the further diffusion of the amylose chains already present in the network. 3. Experimental 3.1. Materials Potato and corn starch of an industrial food grade (NEGRITA) were used to prepare the BC–starch composites. A pure strain of Acetobacter sp. was isolated from a Kombucha tea gel. The initial bacterial strain was incubated in a static culture medium containing 1.0% (w/v) glucose, 1.5% (w/v) peptone, 0.8% (w/ v) yeast extract and 0.3% (w/v) glacial acetic acid, adjusted to pH 3.5 by hydrochloric acid. The culture medium was autoclaved at 121 °C and then allowed to reach room temperature. 0.01% (w/v) cycloheximide and 0.5% (w/v) absolute ethanol were added to prevent the formation of filaments and improve the growth of the cellulose gel. After two days, the strain was extracted, examined by optical microscopy and put into a new culture medium. This process was repeated until the growth of filaments was totally inhibited. The final pure strain was then inoculated and allowed to synthesize cellulose for 21 days at room temperature. 3.2. Preparation of the nanocomposites Starch (corn or potato) was added to the culture medium at concentration of 2.0% (w/v). The starch enriched medium was then autoclaved at 121°. BC–starch gels were collected from the top, washed with water, boiled in 5% (w/v) NaOH for 2400 s, and then washed overnight in 2.5% (w/v) NaOCl. Finally, they were repeatedly rinsed with distilled water. Samples were pressed in a hydraulic plate press (COLLIN) 105 °C to produce sheets. Pure BC and pure starch sheets were also prepared for comparison. 3.3. Methods of characterization 3.3.1. X-ray diffraction (XRD) X-ray diffraction spectra were recorded using a diffractometer (Seifert XRD 3003 TT). Ni-filtered CuKα radiation (wavelength of 0.1542 nm) was produced at 40 kV and 40 mA. Scattered radiation was detected in the angular range of 2.5–40° (2θ) at 0.08° intervals with a
Fig. 1. Scheme of the BC–starch bottom-up process; (a) Starch granules are in suspension in the culture medium; (b) After autoclaving, starch is partially gelatinized, amylose leaches and granules swell; (c) BC nanofibrils grow in presence of the partially gelatinized starch; (d) After hot pressing, the nanocomposite shows interpenetrating networks of amylose and cellulose.
1100
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
Fig. 2. X-ray diffraction spectra of (top to bottom): native corn starch, native potato starch, corn starch sheet, potato starch sheet, BC–corn starch sheet at 2% (w/v), BC–potato starch sheet at 2% (w/v) and pure BC sheet.
measurement time of 5 s per 2θ interval (0.016 2θ per second). The data were analyzed using specialized software (ANALYZE and DRXWin). The d-spacings of the main peaks were calculated according to Bragg's law. The crystallite size was determined using the value of the full-width at half-maximum (FWHM) into the Scherrer Eq. (1): t−
0:89λ β cosθ
ð1Þ
where, λ is the wavelength of the radiation (0.1542 nm), β is the halfwidth at half-maximum (in radians) and θ is the half of the diffraction angle in the plane of analysis. The degree of crystallinity of the samples was estimated following the method of Nara and Komiya [37]. Briefly, a smooth curve connecting the peak baselines is plotted on the diffractograms. The area above the smooth curve (upper diffraction peak area) corresponds to the crystalline portion, while the lower area between the smooth curve and a linear baseline connecting the lower points of the diffraction pattern corresponds to the amorphous section. The degree of crystallinity is estimated by the ratio of the upper diffraction peak area over the total diffraction area. The linear baseline connected two points at 2θ 8.3° and 28° in pure BC samples whereas for BC–starch samples three points at 4.5°, 8.3° and 28° were used to construct the baseline. 3.3.2. Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) ATR-FTIR spectra were recorded with a Fourier transform spectrophotometer (NICOLET NEXUS 470) averaging 64 scans at a resolution of 8 cm− 1 in the range from 4000 to 400 cm− 1. The data were analyzed by means of software (OMNIC). 3.3.3. Mechanical tests Tensile tests were performed in a tensile testing machine (MTS SYNERGIE 200). Rectangular strips (40 mm × 5 mm) of dried samples were cut from the sheets and tested (across 5 samples per composition) with a load cell of 100 N. A crosshead speed of 0.17 mm/s was used. Tensile strength and Young's modulus values were recorded. 3.3.4. Microscopy The qualitative assessment of the morphology of the nanocomposites was carried out by means of Environmental Scanning Electron
Microscopy (ESEM) and Atomic Force Microscopy (AFM). An ESEM XL-30 Philips was used. All samples were sputter coated with gold– palladium and observed using an accelerating voltage of 10 kV. Atomic Force Microscopy (AFM) experiments were performed using a multimode AFM with a NANOTEC head. Samples were viewed applying the tapping mode. Silicon cantilevers with a nominal spring constant of 40 N m− 1 were used. The scan rate ranged 0.25–0.5 Hz and the resonance Frequency was 276 kHz. 4. Results and discussion 4.1. X-ray diffraction The X-ray diffractograms of pure BC and BC–starch nanocomposites are shown in Fig. 2. The native corn starch diffraction peaks (2θ = 15°, 17°, 18° and 23°) are related to its A-type crystalline structure [38] while the peaks observed for potato starch (2θ = 5°, 15°, 17°, 20°, 22° and 24°) reveal its B-type structure [38,39]. Pure starch sheets show lower values of crystallinity, probably associated to the two stages of gelatinization that occur in the process. By contrast, pure BC diffractograms show two well defined peaks in 2θ = 14.5° and 22.6°. The BC–starch nanocomposites diffractograms are similar to those of pure BC, with peaks at 2θ = 5.4°, 14.5° and 22.6° for BC–potato starch sheets and 2θ = 5.6°, 14.5°, 22.7° for BC–corn starch composites. These results suggest that BC forms a coherent nanofibre network in the presence of starch. Amorphous starch is
Table 1 Diffraction peaks, d-spacings, FWHM, crystallite size and degree of crystallinity of pure starch sheets, pure BC sheets and BC–starch composite sheets. 2θ (°) Bacterial cellulose
14.5 19.1 22.6 Bacterial cellulose–potato starch 5.6 (2% w/v) 14.5 22.6 Bacterial cellulose–corn starch 5.4 (2% w/v) 14.5 22.7
d (nm)
FWHM (°)
0.611 1.9208 0.465 1.5794 0.393 1.7344 1.578 17.0200 0.611 2.2359 0.393 1.8860 1.637 3.5318 0.611 2.1393 0.392 1.7947
t Degree of (nm) crystallinity (%) 8.3 10.1 9.2 0.9 7.1 8.5 4.5 7.4 8.9
74.8
69.5
71.0
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
1101
Fig. 3. (a) ATR-FTIR spectra of (top to bottom): corn starch sheet, potato sheet, BC–corn starch sheet at 2% (w/v), BC–potato starch sheet at 2% (w/v), and pure BC; (b) Detail of spectra shown in Fig. 3a, emphasizing the peaks assigned to cellulose Iα and Iβ.
covering the crystalline molecules of cellulose, as confirmed by ESEM images showing cellulose nanofibrils concealed under a starch layer (Fig. 5b,c). The values for d-spacings, crystallite sizes, Full Width at Half Maximum (FWHM) and crystallinity index for BC and BC–starch nanocomposites are shown in Table 1. Pure BC shows the highest degree of crystallinity (74.8%). The presence of amorphous gelatinized starch in the BC–starch nanocomposites accounts for the reduction of crystallinity values to 69.5% in the case of potato starch and 71% with corn starch. The rule of mixtures has been used in the past to estimate the crystallinity of cellulose-pectic polysaccharides solutions using the crystallinities and relative composition of cellulose and pectic
Table 2 Mechanical properties of pure BC, starch, and BC–starch composite sheets.
Fig. 4. Representative strain–stress curves of Pure BC sheet, BC–potato starch nanocomposite sheet and BC–corn starch nanocomposite sheet.
Pure BC Corn starch Potato starch BC–potato starch BC–corn starch
Max. stress (MPa)
Max. elongation (%)
Young's modulus (GPa)
241.42 ± 21.86 19.41 ± 1.90 18.41 ± 1.14 228.94 ± 11.51 206.68 ± 30.18
8.21% ± 3.01 22.45 ± 3.16 12.6 ± 2.20 5.67% ± 1.32 4.79% ± 0.59
6.86 ± 0.32 0.138 ± 0.018 0.285 ± 0.066 6.08 ± 0.73 5.65 ± 0.49
1102
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
polysaccharides [40]. Conversely, in this study, the rule of mixtures has been used to estimate the volume fraction of each component in the nanocomposites according to Eq. (2): kXComposite ¼ kXBC d VBC þ kXStarch d ð1−VBC Þ
ð2Þ
where X is the degree of crystallinity and V is the volume fraction. As shown by the X-ray diffraction spectra (Fig. 2), the crystallinity of pure starch sheets can be considered equal to zero, since no crystallinity peaks were observed in their spectra. Thus, the volume fractions of BC in BC–potato starch nanocomposites and in BC–corn starch nanocomposites, estimated according to Eq. (2), are 92.8% and 95%, respectively.
Fig. 6. ESEM micrograph of (a) Pure BC sheet, (b) BC–potato starch nanocomposite showing the starch covering layer, and some uncovered nanofibrils (white arrows) and (c) BC–corn starch nanocomposite showing the starch covering layer and some uncovered nanofibrils (white arrows).
4.2. Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR)
Fig. 5. AFM micrographs of (a) Pure BC sheet; (b) BC–potato starch nanocomposite sheet and (c) BC–corn starch nanocomposite sheet.
Fig. 3a shows the IR spectra for pure BC and BC–starch composites at 2% (w/v). The spectrum of pure BC shows five characteristic peaks between 984 and 1106 cm− 1 corresponding to the C–O bond stretching mainly attributed to primary alcohols [41]. The band at 1160 cm− 1 is assigned to the C–O–C asymmetric stretching and the peaks at 1313 and 1426 cm− 1 are attributed to CH2 wagging symmetric bending and CH2 symmetric bending, respectively [42]. The bands present between 3200 and 3400 cm− 1 correspond to O–H stretching modes in alcoholic groups. The spectra of the composites are mainly dominated by the cellulose content. It can be observed that for BC–starch composites, there is no presence of the C–O bond stretching at 984 cm− 1 that appears in pure BC.
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
1103
according to AFM images (Fig. 5a). Pure BC samples (Fig. 5a) showed a more defined network structure compared to BC–starch nanocomposites (Fig. 5b,c) as several nanofibres can be observed in the former figure. By contrast, BC–starch nanocomposite sheets (Fig. 5b,c) did not show a network of nanofibres. The qualitative differences in morphology of pure BC and BC–starch sheets can be appreciated in the ESEM micrographs (Fig. 6). The usual morphology of a BC sheet is shown in Fig. 6a, while Fig. 6b,c show the starch layer that covers the fibrils network of BC–starch nanocomposite sheets. This layer seems quasi homogeneous at the meso- and microlevels. The lack of regularity of the starch layer can be seen in Fig. 6b,c, where regions of covered fibrils are near regions of partially covered and uncovered fibrils. Furthermore, in Fig. 6c it can be observed (white arrows) that one single fibre has covered and uncovered parts along its length. The surface morphology of cryo-fractured specimens is shown in Fig. 7. Pure BC (Fig. 7a) and BC–corn starch sheets are formed by a pile of thin layers. This morphology is typical for paper and other pulp based products and in the case of these BC composites, it was produced by the hot pressing process. The coherent network seems to be disrupted in the fracture surface as revealed by some pulled out fibres. The fact that pure BC and BC–starch nanocomposites have a similar morphology at low magnifications confirms that the aggregation of starch takes place at the nanolevel. 5. Conclusions
Fig. 7. ESEM micrograph of cryo-fractured surfaces of (a) Pure BC sheet, (b) BC–corn starch nanocomposite sheet.
Also, for pure BC two shoulders can be observed in the region between 700 and 760 cm− 1 (Fig. 3b). These are attributed to the Iα and Iβ crystalline forms which are typical of bacterial cellulose [18,43]. Furthermore, these crystalline forms remain in BC–starch composites but with a decreased intensity, being the Iβ form more noticeable and indicating that there might be interaction between the amorphous starch, determined by X-ray diffraction, and the predominant cellulose.
Bacterial cellulose–starch self-assembled nanocomposites were prepared by allowing the bottom-up formation of a coherent network of cellulose nanofibres in the presence of partially gelatinized starch granules. This bottom-up technique preserved the typical network of cellulose fibres as there was no need to disintegrate the BC gel in order to combine it with a second phase. By contrast, a coherent nanostructure that can be controlled and varied at different stages was obtained. The bottom-up technique seemed to be an adequate approach to manufacture BC nanocomposites. Starch was gelatinized in different stages and formed a layer that covers the cellulose nanofibrils. The volume fraction of the strong phase in the nanocomposites produced by this technique was around 90%, which is similar to the proportions present in some biological nanocomposites, such as nacre. Structural properties determined by XRD and ATR-FTIR showed that the crystallinity of BC was preserved in spite of the presence of starch, hence the mechanical properties of the nanocomposites showed no significant decrease.
4.3. Mechanical properties Acknowledgments Fig. 4 shows representative stress–strain curves. As it is confirmed by the values presented in Table 2, the BC–starch nanocomposites showed no significant decrease in their mechanical properties with regard to pure BC, assessed by tensile tests. Stress at break and Young's modulus values are similar for every sheet tested. Maximum stress values range from 200 to 250 MPa, while Young's modulus values were between 5.6 and 6.8 GPa (Table 2). Strain at break values of pure BC sheets were around 8% whereas the nanocomposite sheets failed at slightly lower strain levels (4.5–6%). Overall, the lower mechanical properties were obtained for samples with corn starch, where samples with potato starch displayed a more similar behavior with regard to pure BC. This confirms that the good mechanical properties of BC networks are kept almost intact by using the bottom-up technique described here. Data reported in the literature for BC nanocomposites produced by a standard blending technique [44] show important decreases in mechanical properties with regard to pure BC. 4.4. BC nanocomposites morphology The morphology of pure BC sheets is well known and has been reported elsewhere as a coherent network of interconnected nanofibrils [3–5]. The diameters of these nanofibrils are in the range 100–150 nm
The authors would like to thank the Direction of Research (DAI) of PUCP, the International Foundation for Science (IFS, Stockholm, Sweden; RGA F/4194-1) and the Generalitat Valenciana, Conselleria de Empresa, Universidad y Ciencia (Project number ARVIV/2007/101) for financial support. CJG thanks the Oficina de Relaciones Internacionales of the Universitat de Valencia for financial support. The authors gratefully acknowledge Dr. M. Carmen Bano from the Departament de Bioquimica y Biologia Molecular of the Universitat de Valencia and Prof. Dora Maurtua from the Department of Microbiology of Universidad Cayetano Heredia for the provided facilities and assistance with the isolation of Acetobacter. Dr. Daniel Lopez from the ICTP-CSIC is also acknowledged. References [1] [2] [3] [4] [5] [6] [7]
J. Elemans, A.E. Rowan, R.I.M. Nolte, J. Mater. Chem. 13 (2003) 2661. S.G. Zhang, Nat. Biotechnol. 21 (2003) 1171. A. Hirai, M. Tsuji, F. Horii, Cellulose 9 (2002) 105. M. Iguchi, S. Yamanaka, A. Budhiono, J. Mat. Sci. 35 (2000) 261. A.N. Nakagaito, S. Iwamoto, H. Yano, App. Phys. A 80 (2005) 93. W. Gindl, J. Keckes, Comp. Sci. Tech. 64 (2004) 2407. H. Yano, J. Sugiyama, A.N. Nakagaito, M. Nogi, T. Matsuura, M. Hikita, K. Handa, Adv. Mat. 17 (2005) 153.
1104
C.J. Grande et al. / Materials Science and Engineering C 29 (2009) 1098–1104
[8] J.D. Fontana, A.M. de Sousa, C.K. Fontana, I.L. Torriani, J.C. Moreschi, B.J. Gallotti, S.J. de Sousa, G.P. Narcisco, J.A. Bichara, L.F. Farah, Appl. Biochem. Biotechnol. 24 (1990) 253. [9] A.B. Novaes Jr, A.B. Novaes, Clin. Oral Implant Res. 4 (1993) 106. [10] B. dos Anjos, A.B. Novaes Jr, R. Meffert, E.P.J. Barboza, J. Periodontol. 69 (1998) 454. [11] G. Helenius, H. Backdahl, A. Bodin, U. Nannmark, P. Gatenholm, B. Risberg, J. Biomed. Mater. Res. A 76 (2006) 431. [12] K. Watanabe, M. Tabuchi, Y. Morinaga, F.A. Yoshinaga, Cellulose 5 (1998) 187. [13] H. Bäckdahl, G. Helenius, A. Bodin, U. Nannmark, B.R. Johansson, B. Risberg, P. Gatenholm, Biomaterials 27 (2005) 2141. [14] A. Svensson, E. Nicklasson, T. Harrah, B. Panilaitis, D.L. Kaplan, M. Brittberg, P. Gatenholm, Biomaterials 26 (2005) 419. [15] C.R. Rambo, D.O.S. Recouvreux, C.A. Carminatti, A.K. Pitlovanciv, R.V. Antônio, L.M. Porto, Mat. Sci. Eng. C 24 (2008) 549. [16] S. Bae, M. Shoda, Biotech. Bioeng. 90 (2005) 20. [17] S. Bae, Y. Sugano, M. Shoda, J. Biosci. Bioeng. 97 (2004) 33. [18] R.H. Atalla, D.L. VanderHart, Science 223 (1984) 283. [19] S. Yamanaka, M. Ishihara, J. Sugiyama, Cellulose 7 (2000) 213. [20] A. Bohn, H.P. Fink, J. Ganster, M. Pinnow, Macromol. Chem. Phys. 201 (2000) 1913. [21] H.S. Barud, C. Barrios, T. Regiani, R.F.C. Marques, M. Verelst, J. Dexpert-Ghys, Y. Messaddeq, S.J.L. Ribeiro, Mat. Sci. Eng. C 28 (2008) 515. [22] Z. Wan, Y. Huang, C.D. Yuan, S. Raman, Y. Zhu, H.J. Jiang, F. Hea, C. Gao, Mat. Sci. Eng. C 27 (2007) 855. [23] J. Touzel, B. Chabbert, B. Monties, P. Debeire, B. Cathala, J. Agric. Food Chem. 51 (2003) 981. [24] R. Mormino, H. Bungay, Appl. Microbiol. Biotechnol. 62 (2003) 503. [25] S.C. Mendes, R.L. Reis, Y.P. Bovell, A.M. Cunha, C.A. van Blitterswijk, J.D. de Bruijn, Biomaterials 22 (2001) 2057.
[26] J. Nakamatsu, F.G. Torres, O.P. Troncoso, Y.M. Lin, A.R. Boccaccini, Biomacromolecules 7 (2006) 3345. [27] F.G. Torres, A.R. Boccaccini, O.P. Troncoso, J. App. Polym. Sci. 103 (2007) 1332. [28] C.J. Grande, F.G. Torres, C.M. Gomez, Composites of bacterial cellulose–hydroxyapatite for biomedical applications. Advances in Polymeric Materials, Spanish Polymer Group (GEP) 10th Meeting, September 16–20, 2007, Sevilla, Spain. [29] P.J. Jenkins, A.M. Donald, Carbohydr. Res. 308 (1993) 133. [30] R. Vermeylen, V. Derycke, J.A. Delcour, B. Goderis, H. Reynaers, M.H.J. Koch, Biomacromolecules 7 (2006) 2624. [31] S.L. Randzio, I. Flis-Kabulska, J.E. Grolier, Macromolecules 35 (2002) 8852. [32] M. Schramm, Z. Gromet, S. Hestrin, Biochem. J. 67 (1957) 669. [33] E.R. Forng, S.M. Anderson, R.E. Cannon, Appl. Environ. Microbiol. 55 (1989) 1317. [34] C.H. Haigler, A.R. White, R.M. Brown, K.M. Cooper, The Journal of Cell Biology 94 (1982) 64. [35] Q. Zhang, Z. Lu, H. Hu, W. Yang, P.E. Marszalek, J. Am. Chem. Soc. 128 (2006) 9387. [36] A.G.J. Kuipers, E. Jacobsen, R.G.F. Visser, The Plant Cell 6 (1994) 43. [37] S. Nara, T. Komiya, Starch 35 (1983) 407. [38] N.W.H. Cheetham, L. Tao, Carbohydr. Polym. 36 (1998) 277. [39] A. Rindlav, S.H.D. Hulleman, P. Gatenholm, Carbohydr. Polym. 34 (1997) 25. [40] D.M.R. Georget, P. Cairns, A.C. Smith, K.W. Waldron, Int. J. Bio. Macromol. 26 (1999) 325. [41] Y. Marechal, H. Chanzy, J. Mol. Struc. 523 (2000) 183. [42] M. Kacurakova, A.C. Smith, M.J. Gidley, R.H. Wilson, Carbohydr. Res. 337 (2002) 1145. [43] W. Czaja, D. Romanovicz, R.M. Brown Jr., Cellulose 11 (2004) 403. [44] S. Yamanaka, K. Watanabe, N. Kitamura, J. Mat. Sci. 24 (1989) 3141.
