Journal of Colloid and Interface Science 377 (2012) 8893

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Conductive properties of TiO2/bacterial cellulose hybrid bres

Junkal Gutierrez, Agnieszka Tercjak , Itxaso Algar, Aloa Retegi, Iaki Mondragon
Materials + Technologies Group, Dpto. Ingeniera Qumica y M. Ambiente, Escuela Politcnica, Universidad Pas Vasco/Euskal Herriko Unibertsitatea, Pza. Europa 1, 20018 Donostia-San Sebastin, Spain

a r t i c l e

i n f o

a b s t r a c t
Conductive properties of TiO2 nanoparticles and TiO2/BC hybrid inorganic/organic bres were investigated by electrostatic force microscopy (EFM). TiO2/BC hybrid composites were prepared based on bacterial cellulose produced by Gluconobacter xylinum, being the bacterial cellulose as a hydrophilic substrate for TiO2 nanoparticles synthesized via solgel. Taken into account hydrophilic nature of the cellulose, TiO2 nanoparticles were located on the surface of the bres due to hydrogen bonding interactions. EFM was used to determine qualitatively conductive properties of TiO2 nanoparticles and their TiO2/BC hybrid inorganic/organic bres. Results indicate that TiO2/BC hybrid bres respond to applied bias regardless of the sign of the applied voltage. 2012 Published by Elsevier Inc.

Article history: Received 19 December 2011 Accepted 24 March 2012 Available online 7 April 2012 Keywords: Bacterial cellulose Hybrid bres TiO2 nanoparticles Solgel Electrostatic force microscopy

1. Introduction Nowadays new research area focused on hybrid inorganic/organic bres materials based on bacterial cellulose and inorganic nanoparticles is growing [113]. Bacterial cellulose (BC) is an attractive kind of cellulose produced by acetic acid bacteria, Gluconobacter xylinum [14,15]. BC has a diversity of excellent properties such as good biocompatibility, high tensile strength, high water retention capability, high hydrophilicity and high crystallinity [16,17]. These properties put forward bacterial cellulose as a perfect hydrophilic substrate. Several works have been reported on titanium dioxide/bacterial cellulose (TiO2/BC) hybrid inorganic/organic bres [913]. TiO2 nanoparticles are of particular interest taking into account their properties, which offer wide range of application in electric and photocatalytic elds [1820]. The most common method for the generation of TiO2 nanoparticles is the solgel synthesis based on titanium alkoxides. This solgel method leads to the well-dened nanoparticles with uniform size. This method allows the preparation of hybrid inorganic/organic materials due to the in situ generation of inorganic particles, which are uniformly dispersed at the nanometric scale due to hydrogen or covalent bonds [8]. Thus, using titanium isopropoxide as solgel precursor, inorganic network having hydroxyl groups on the surface can be generated. These inorganic networks surrounded by hydroxyl groups can interact with hydrophilic bacterial cellulose that acts as substrate and consequently, TiO2/BC hybrid bres can be obtained.
Corresponding authors.
E-mail addresses: agnieszka.tercjaks@ehu.es (A. Tercjak), inaki.mondragon@ ehu.es (I. Mondragon). 0021-9797/$ - see front matter 2012 Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jcis.2012.03.075

The research work regarding bacterial cellulose with TiO2 nanoparticles has been focused on the fabrication conditions and their inuence on the nal morphology and properties of the hybrid inorganic/organic bres [913]. These inorganic/organic composite is of special interest reaching novel characteristics corresponding to both inorganic and organic parts and even novel properties. Hybridization of the organic and inorganic components allows to design multifunctional materials based on bacterial cellulose with interesting conductive, optical and mechanical properties, which can nd wide range of application in nanotechnology [113]. However, to the best of our knowledge, conductive properties of TiO2/BC hybrid composites have not been yet reported. Taken above into account, in this study, conductive properties of TiO2 nanoparticles and TiO2/BC hybrid bres have been investigated by means of electrostatic force microscopy (EFM). EFM allows to measure qualitatively the electric eld gradient distribution above the sample surface. The EFM probe performs simultaneously two scans, the rst one to obtain the topography in tapping mode and the second at a chosen lift height to measure the electrostatic force [2124]. Different voltages have been applied to check conductive properties of designed systems. Additionally, the inuence of the sign (positive or negative) of the applied voltage on the response of the samples has been also investigated. 2. Materials and methods 2.1. Bacterial cellulose preparation Bacterial cellulose was obtained in our laboratory from a G. xylinum pellicle incubated for 13 days at 28 C in a static culture following the preparation method published by Retegi et al. [25].

J. Gutierrez et al. / Journal of Colloid and Interface Science 377 (2012) 8893

89

Fig. 1. AFM phase images (5 lm 5 lm) of (A) TiO2 solgel cast solution before, (B) after thermal treatment and (C) XRD patterns obtained nanoparticles.

2.2. TiO2 solgel Solgel solution used to obtain TiO2 nanoparticles was carried out with titanium isopropoxide (Ti(OCH(CH3)2)4), TTIP) from Aldrich; isopropanol (IPA), hydrochloric acid (HCl, 37%) and toluene were of analytical grade. The solgel solution was prepared by mixing IPA (5 mL), TTIP (0.125 mmol), HCl (0.125 mmol) and toluene (5 mL) under vigorous stirring during 1 h [26,27]. TiO2 solgel thin lms were prepared by spin-coating onto Si (1 0 0) wafers at 2000 rpm using a Specialty Coating Systems Inc. spin-coater.

3. Results and discussion TiO2 nanoparticles were synthesized using solgel technique. Fig. 1 shows surface morphology of TiO2 solgel lm as-cast (Fig. 1A) and after heat treatment at 500 C for 3 h (Fig. 1B). In Fig. 1A, AFM phase image before heat treatment shows a continuous coating that cover Si (1 0 0) wafer surface. Obtained morphology was smooth and no differences on the lm surface were detected. In the case of Fig. 1B, AFM phase image after heat treatment showed highly dense arrays of uniformly sized TiO2 nanoparticles. Additionally, the evolution of the crystalline structure of the TiO2 nanoparticles before and after heat treatment was observed by X-ray diffractometry (Fig. 1C). TiO2 nanoparticles lm before heat treatment showed only very broad and diffuse patterns, which indicate their characteristic amorphous nature. On the contrary, the XRD pattern of the TiO2 nanoparticles after heat treatment became intense and narrow conrming the crystallization of the TiO2 nanoparticles. Deeper XRD analysis allows to deduce that the heat treatment resulted in diffraction peaks related to both anatase and rutile phases [28] of TiO2 (63/37 wt.%). In order to clarify the morphology of generated TiO2 nanoparticles, before and after thermal treatment, high magnication images of Fig. 1 were performed (Fig. 2). Fig. 2A shows non-uniform nanoparticles with irregular shapes and an average size of 10 nm. On the contrary, as can be observed in Fig. 2B, heat treatment allows to design very regular, spherical TiO2 nanoparticles homogenously deposited on the Si (1 0 0) wafer surface. It can be clearly detected that the average size of the TiO2 nanoparticles was $25 nm in diameter. The thermal treatment provokes an increase in size and crystallization of nanoparticles. Representative AFM phase images of bacterial cellulose bres used as hydrophilic substrate for the preparation of the TiO2/BC composite are shown in Fig. 3A and B. In Fig. 3A, BC bres can be easily observed. The width of homogeneous BC bres deposited on the Si (1 0 0) wafer surface was between 150 and 200 nm. The bres length cannot be determined taking into account that bres-ends are not apparent in the analysed AFM phase image. Nevertheless, in average, the bres length is higher than 15 lm. Fig. 3B shows AFM high magnication images (height and phase) of an individual BC bre. Deeper analysis of these images allows to clearly detect the typical conguration of the individual microbrils inside the bres; each bre consists of 35 microbrils held by electrostatic forces [29]. The microbrils detection was conrmed by AFM height prole (top left inset of Fig. 3B), where individual BC bre consisting of several microbrils can be distinguished. The width of the individual microbrils was $25 nm as indicated by the prole. Here, it should be pointed out that individual BC microbrils had smaller thickness than

2.3. TiO2/BC hybrid bres First, 0.1 wt.% BC in isopropanol was sonicated at room temperature by using a microprocessor sonicator 750 W (Vibracell 75043) from Bioblock Scientic with an amplitude range between 20% and 25%. After 1 h, 25 vol.% solgel was added and sonicated for an additional 30 min. Finally, the prepared suspension was ltered using a 0.2 lm PTFE lter.

2.4. Methods Atomic force microscopy (AFM) and electrostatic force microscopy images were obtained in tapping mode using a Dimension 3100 NanoScope IV (Veeco, USA). Etched single beam cantilever (110140 lm length) silicon nitride probes having a tip nominal radius of 10 nm were used. Scan rates were ranged from 0.7 to 1.2 Hz s1. Measurements were performed with 512 scan lines and target amplitude around 0.9 V. In order to obtain repeatable results of TiO2 nanoparticles, BC and TiO2/BC hybrid bres morphology, different regions of the lms were scanned. Electrostatic force microscopy was used to study the conductive capability of TiO2 nanoparticles and their composites with BC. These measurements were performed using the same scanning probe microscope operating in the lift-mode (lift height was high enough to avoid topographic effect) in ambient conditions and equipped with an integrated Co/Cr-coated MESP tip having a resonance frequency around 75 kHz. The secondary imaging mode derived from the tapping mode that measures the electric eld gradient distribution above the sample surface was detected by applying a voltage to the cantilever tip. Thermogravimetric analysis was performed with a TGA/SDTA851e equipment under air atmosphere at a heating rate of 10 C min1 from room temperature to 800 C. X-ray diffraction patterns have been measured using a X-ray diffractometer PANalytical Xpert PRO. Samples were scanned from 10 to 70 2h.

90

J. Gutierrez et al. / Journal of Colloid and Interface Science 377 (2012) 8893

Fig. 3. AFM phase image of (A) bacterial cellulose bres and (B) AFM high magnication images (height and phase). The inset corresponds to the prole of bre cross-section.

Fig. 2. High magnication AFM phase images (500 nm 500 nm) of (A) TiO2 sol gel cast solution before and (B) after thermal treatment.

width [30,31], which occasionally leads the brils to twist over their axis mostly in the direction of their smaller thickness [29]. The possible physical interactions between BC bres and sol gel solution in the TiO2/BC hybrid bres were shown in Scheme 1. As it is well known, hydroxyl groups are always present on the

surface of the BC bres and provoke the H-bonding interaction between BC bres and hydrophilic inorganic TiO2 solgel network. In other words, taken into account that the free OH are present, generated solgel TiO2 nanoparticles can be located on the BC bres surface. The digital image of the synthesized TiO2/BC composites is shown in Fig. 4A. After nal ltration step, designed material was a continuous white colour lm attributed to the presence of TiO2 nanoparticles. Fig. 4B shows TGA curves carried out in static air obtained for TiO2 solgel, BC and TiO2/BC hybrid bres. As can be observed, for BC bres, three weight losses were detected. The rst slight weight loss from ambient to 75 C can be attributed to BC

Scheme 1. Illustration of the synthesis of TiO2/BC composites.

J. Gutierrez et al. / Journal of Colloid and Interface Science 377 (2012) 8893

91

Fig. 5. (A) AFM phase image (2 lm 2 lm) of TiO2/BC composite and (B) high magnication of part a.

Fig. 4. (A) Digital image of TiO2/BC composites, (B) TGA curves of TiO2 solgel, BC and TiO2/BC composite and (C) XRD patterns of BC and TiO2/BC composite.

dehydration [1,2], and the other two steps correspond to the thermal degradation that occurred at 340 C and 515 C [1,2,32]. Taking into account that TiO2/BC composite was prepared via solgel method, the rst weight loss can be attributed to physically adsorbed and hydrogen bonded linked solvent molecules, water and isopropanol. This behaviour can be conrmed by the TGA curve of the TiO2 solgel nanoparticles where only one weight loss was detected. Two other weight loss steps can be related to the decomposition of BC bres. It should be note that the decomposition temperature of TiO2/BC composites (280 C) was lower than in the case of BC (330 C). This phenomenon can be related to the catalytic property of incorporated TiO2 nanoparticles [12,32]. Addi-

tionally, it is observed that in the case of TiO2/BC thermogram, there is no weight loss after the sample reached 450 C. The residual mass was used to calculate the inorganic content in the synthesized composite taking into account that BC matrix did not generate any char. Thus, for synthesized TiO2/BC composite, the content of TiO2 nanoparticles was 40 wt.%. In order to analyse the inuence of the incorporation of TiO2 nanoparticles on the crystallinity of BC bres, X-ray diffraction was used. XRD patterns of BC and TiO2/BC hybrid bres are shown in Fig. 4C. BC bres displayed the typical XRD pattern of cellulose I with the main diffraction signals at around 2h = 14.6, 16.8, 22.8 and 34.9, assigned to the (1 0 1), (1 0 1), (0 0 2) and (0 4 0) diffraction planes, respectively [33,34]. In the case of TiO2/BC composite, the main characteristic peaks correspond to the typical prole of cellulose I were also clearly distinguished; however, this pattern reveals a noticeable decrease in the crystallinity if compared with BC bres pattern. This can be related with the use of solgel process for the generation of TiO2 nanoparticles since as mentioned above they were in amorphous state without any additional heat treatment. Thus, the decreasing of the crystallinity in the composite was related to the incorporation of the amorphous TiO2 nanoparticles into BC bres matrix.

92

J. Gutierrez et al. / Journal of Colloid and Interface Science 377 (2012) 8893

AFM phase image of synthesized TiO2/BC hybrid bres is shown in Fig. 5. Under the same preparation conditions, the surface of the modied bacterial cellulose bres appeared rougher, in the AFM image, if compared to the surface of neat BC bres (see Fig. 3). The width of TiO2/BC hybrid bres is around 80 nm. In order to verify the changes on the surface of the modied bres and to visualize TiO2 nanoparticles, a high magnication image is shown as inset at the bottom right of Fig. 5. Small bright spots were clearly observed covering bres surface, which means that solgel synthesis allowed to deposit TiO2 nanoparticles on the surface of the BC bres. Moreover, no aggregation of TiO2 nanoparticles on the surface of bacterial cellulose bres was detected, thus conrming good distribution of TiO2 nanoparticles on BC bres surface. The size of the TiO2 nanoparticles located on the bres surface was around 510 nm in diameter. Moreover, AFM phase image conrms the fact that BC bres can act as hydrophilic substrate for the generation of TiO2 spherical nanoparticles generated via sol gel. From the point of view of possible applications of the synthesized TiO2/BC composite, conductive properties were measured using EFM technique, which allowed to qualitatively detect charge and uncharged domains applying bias voltage into EFM-tip. In order to investigate the conductive behaviour of spherical pristine TiO2 nanoparticles, EFM measurements of TiO2 nanoparticles thin lms before and after heat treatment were performed (Fig. 6). For both samples initially, 0 bias voltage was applied in order to conrm the non-inuence of the topographic effect. In the case of EFM measurements of TiO2 cast solution before thermal treatment (Fig. 6A and B), no charge domains were detected on the surface regardless of both the value and the sign of the applied bias voltage. Taking this into account, one can conclude that under the measurement conditions, TiO2 cast solution before thermal treatment does not respond to the applied voltage. Thus, under this preparation conditions, synthesized solgel coat was electrically neutral. Fig. 6C shows EFM phase image of TiO2 nanoparticles after thermal treatment by applying different positive bias voltage. As can be observed in corresponding 0 bias voltage EFM phase image, no charge domains were detected on the lm surface. On the contrary,

when positive bias voltage was applied, TiO2 nanoparticles appeared as bright dots on the darker silicon wafer surface. Taking into account that voltages applied to the conductive EFM-tip allow to detect only charged zones in the EFM phase image, obtained results clearly indicated that spherical TiO2 nanoparticles respond to the applied voltage. Moreover, it should be mentioned that increasing of the applied voltage resulted in increasing of EFM phase image contrast between charged and uncharged zones, which indicated strong inuence of the applied bias voltage on the investigated TiO2 nanoparticles lm. Similar behaviour was also observed in the case of applied negative bias voltage. In Fig. 6D, EFM phase image of spherical TiO2 nanoparticles with different applied negative bias (0, 3 and 9 V) is shown. Obtained results conrm that spherical TiO2 nanoparticles also respond to the negative bias voltage. Increasing of negative voltage value provoked higher contrast between charged and uncharged zones in EFM phase image. Applying the same positive or negative bias value means almost the same contrast in EFM phase image (Fig. 6C and D). Consequently, one can conclude that both positive and negative bias provoke almost the same spherical TiO2 nanoparticles charge response. Fig. 7 shows EFM phase images of TiO2/BC composite applying different bias voltages. Here, it should be noted that the EFM image of TiO2/BC composites is different if compared to the AFM phase image (Fig. 5) due to the different resolution and coating of the tip used for EFM measurement. In the case of AFM phase image, tip resolution allows the detection of the small individual TiO2 nanoparticles. On the contrary, the resolution of the EFM tip was able to detect conductive areas due to the small size (510 nm in diameter) of TiO2 nanoparticles. As shown in Fig. 7A, EFM phase images of synthesized TiO2/BC hybrid bres were performed applying 3, 0 and 3 V bias voltage to the conductive EFM-tip. As well as for TiO2 nanoparticles thin lm, also in this case, a bias of 0 V was applied to the tip in order to conrm the non-inuence of the topographic effect. It is worth to note that the same positive and negative biases provoked almost the same contrast in EFM phase image. The inuence of the bias voltage value can be observed in Fig. 7B, where 9, 0 and 9 bias voltages were applied. The increase in the applied bias voltage value leads to higher EFM phase image contrast. Thus, these measurements allow to conclude that applying the same level of positive or negative bias to the EFM tip results in almost the same contrast in the EFM phase image. Moreover, no differences were observed when the same bias value switched from negative to positive, indicating that investigated TiO2/BC composite was able to respond, regardless of the sign of the applied voltage. Here, it should be pointed out that the fact that the amorphous TiO2 nanoparticles conned to cellulose bres respond to the voltage applied to the EFM-tip can be related to the strong electrostatic dipoledipole interactions generated by hydrogen bonding

Fig. 6. EFM phase images (3 lm 3 lm) of (A and B) TiO2 solgel cast solution before and (C and D) after thermal treatment.

Fig. 7. EFM phase images (3 lm 3 lm) of TiO2/BC composite lm applying (A) 3, 0 and 3 and (B) 9, 0 and 9 bias voltages.

J. Gutierrez et al. / Journal of Colloid and Interface Science 377 (2012) 8893

93

interactions between the hydroxyl groups in the cellulose, and the hydroxyl groups on the inorganic oxide provoke high electric dipole moment of TiO2 nanoparticles in the TiO2/BC hybrid bres. Additionally, it should be mentioned that the conductive or nonconductive behaviour of the solgel coat without thermal treatment depends strongly on the preparation conditions. 4. Conclusions TiO2/BC hybrid inorganic/organic bres were synthesized using solgel method for the generation of the TiO2 nanoparticles. BC bres act as hydrophilic substrate, and TiO2 nanoparticles were located on their surface due to hydrogen bonding interactions. Atomic force microscopy images allow to investigate the morphology of the synthesized TiO2 nanoparticles and TiO2/BC hybrid bres. Additionally, electrostatic force microscopy was used to study the conductive properties of TiO2 nanoparticles. A comparison between EFM measurements of amorphous neat TiO2 nanoparticles and amorphous TiO2 nanoparticles located on the surface of bacterial cellulose bres reveals the inuence of the electrostatic dipoledipole interactions generated between bacterial cellulose and TiO2 nanoparticles on the conductive behaviour. Amorphous neat TiO2 nanoparticles were electrically neutral and consequently do not respond to the applied voltage. On the contrary, in the case of TiO2/BC hybrid bres, obtained results indicate that TiO2 nanoparticles are able to responds on the voltage applied to the EFM tip regardless of the sign of the applied voltage. Moreover, increasing applied voltage value leads to higher contrast in the EFM phase images, suggesting that increase in applied voltage provokes higher response of the TiO2 located on the surface of the bacterial cellulose bres. Acknowledgments Financial support from the Basque Country Government in the frame of Grupos Consolidados (IT-365-07), NANOTES (SPE10UN40) and from the Spanish Ministry of Science and Innovation (MAT2009-12832) is gratefully acknowledged. Additionally, J.G. thanks Eusko Jaurlaritza/Gobierno Vasco (Programas de becas para formacin y perfeccionamiento de personal investigador) and A.T. acknowledges MICINN for Ramn y Cajal program. Moreover, we are grateful to the Macrobehavior Mesostructure and X-ray Service: Analysis of Rocks and Minerals Nanotechnology SGIker units of the UPV/EHU. Unit

References
[1] H.S. Barud, R.M.N. Assuno, M.A.U. Martines, J. Dexpert-Ghys, R.F.C. Marques, Y. Messaddeq, S.J.L. Ribeiro, J. SolGel Sci. Technol. 46 (2008) 363. [2] X. Li, S. Chen, W. Hu, S. Shi, W. Shen, X. Zhang, H. Wang, Carbohydr. Polym. 76 (2009) 509. [3] S. Ifuku, M. Tsuji, M. Morimoto, H. Saimoto, H. Yano, Biomacromolecules 10 (2009) 2714. [4] R.J.B. Pinto, P.A.A.P. Marques, A.M. Barros-Timmons, T. Trindade, C.P. Neto, Compos. Sci. Technol. 68 (2008) 1088. [5] A. Hou, Y. Yu, H. Chem, Carbohydr. Polym. 79 (2010) 578. [6] S. Yano, H. Maeda, M. Nakajima, T. Hagiwara, T. Sawaguchi, Cellulose 15 (2008) 111. [7] P. Kulpinski, J. Appl. Polym. Sci. 98 (2005) 1793. [8] A. Hou, Y. Shi, Y. Yu, Carbohydr. Polym. 77 (2009) 201. [9] P.A.A.P. Marques, T. Trindade, C.P. Neto, Compos. Sci Technol. 66 (2006) 1038. [10] G. Goutailler, C. Guillard, S. Daniele, L.G. Hubert-Pfalzgraf, J. Mater. Chem. 13 (13) (2003) 342. [11] D. Sun, J. Yang, X. Wang, Nanoscale 2 (2010) 287. [12] J. Zeng, S. Liu, J. Cai, L. Zhang, J. Phys. Chem. C 114 (2010) 7806. [13] W.A. Daoud, J.H. Xin, Y.-H. Zhang, Surf. Sci. 599 (2005) 69. [14] D. Klemm, B. Heublein, H.P. Fink, A. Bohn, Angew. Chem., Int. 22 (2005) 3358. [15] P. Ross, R. Mayer, M. Benziman, Microbiol. Rev. 55 (1991) 35. [16] D. Klemm, D. Schumann, U. Udhardt, S. Marsch, Prog. Polym. Sci. 26 (2001) 1561. [17] A. Svensson, E. Nicklasson, T. Harrah, B. Panilaitis, D.L. Kaplan, M. Brittberg, P. Gatenholm, Biomaterials 26 (2005) 419. [18] M. Koelsch, S. Cassaignon, C. Minh, J.F. Guillemoles, J.P. Jolivet, Thin Solid Films 86 (2004) 451. [19] Y.V. Kolenko, B.R. Churagulov, M. Kunst, L. Mazerolles, C.C. Justin, Appl. Catal., B 8 (2004) 54. [20] Y. Li, T.J. White, S.H. Lim, J. Solid State Chem. 177 (2004) 1372. [21] L. Peponi, A. Tercjak, J. Gutierrez, M. Cardinali, I. Mondragon, L. Valentini, J.M. Kenny, Carbon 48 (48) (2010) 2590. [22] A. Tercjak, I. Garcia, I. Mondragon, Nanotechnology 19 (2008) 275701. [23] J. Gutierrez, A. Tercjak, L. Peponi, I. Mondragon, J. Phys. Chem. C 113 (2009) 8601. [24] A. Tercjak, J. Gutierrez, G. Mondragon, I. Mondragon, J. Phys. Chem. C 115 (2011) 22180. [25] A. Retegi, N. Gabilondo, C. Pea, R. Zuluaga, C. Castro, P. Gaan, K. de la Caba, I. Mondragon, Cellulose 17 (2010) 661. [26] Z. Sun, D.H. Kim, M. Wolkenhauer, G.G. Bumbu, W. Knoll, J.S. Gutmann, ChemPhysChem 7 (2006) 370. [27] Y.J. Cheng, J.S. Gutmann, J. Am. Chem. Soc. 128 (2006) 4658. [28] S. Bakardjievaa, J. Subrta, V. Stengla, M.J. Dianezb, M.J. Sayaguesb, Appl. Catal., B 58 (2005) 193. [29] R.T. Olsson, R. Kraemer, A. Lopez-Rubio, S. Torres-Giner, M.J. Ocio, J.M. Lagaron, Macromolecules 43 (2010) 4201. [30] I. Ohad, D.J. Danon, Cell Biol. 22 (1964) 302. [31] I. Ohad, D.J. Mejzler, J. Polym. Sci., Part A: Polym. Chem. 3 (1965) 399. [32] H.S. Barud, C.A. Ribeiro, M.S. Crespi, M.A.U. Martines, G.H.Y.S. Dexpert, J. Marques, F.C. Rodrigo, Y. Messaddeq, S.J.L. Ribeiro, J. Therm. Anal. Calorim. 87 (2007) 815. [33] W. Czaja, D. Romanovicz, R.M. Brown, Cellulose 11 (2004) 403. [34] G. Meshitsuka, A. Isogai, Chemical Structures of Cellulose, Hemicelluloses, and Lignin, CRC Press, New York, 1995. p. 11.