Elastic Modulus of Single Cellulose Microfibrils from Tunicate

Biomacromolecules 2009, 10, 2571–2576

2571

Elastic Modulus of Single Cellulose Microfibrils from Tunicate Measured by Atomic Force Microscopy Shinichiro Iwamoto, Weihua Kai, Akira Isogai, and Tadahisa Iwata* Department of Biomaterial Sciences, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan Received May 6, 2009; Revised Manuscript Received July 13, 2009

The elastic modulus of single microfibrils from tunicate (Halocynthia papillosa) cellulose was measured by atomic force microscopy (AFM). Microfibrils with cross-sectional dimensions 8 × 20 nm and several micrometers in length were obtained by oxidation of cellulose with 2,2,6,6-tetramethylpiperidine-1-oxyl radical (TEMPO) as a catalyst and subsequent mechanical disintegration in water and by sulfuric acid hydrolysis. The nanocellulosic materials were deposited on a specially designed silicon wafer with grooves 227 nm in width, and a three-point bending test was applied to determine the elastic modulus using an AFM cantilever. The elastic moduli of single microfibrils prepared by TEMPO-oxidation and acid hydrolysis were 145.2 ( 31.3 and 150.7 ( 28.8 GPa, respectively. The result showed that the experimentally determined modulus of the highly crystalline tunicate microfibrils was in agreement with the elastic modulus of native cellulose crystals.

Introduction Cellulose is a linear natural polymer consisting of Danhydroglucose units joined by (1f4)-β-glycosidic linkages. Plants, fungi, and algae as well as animals can produce cellulose. It is synthesized from membrane complexes, generally referred to as terminal complexes (TCs). Cellulose aggregates regularly along the chain, resulting in intermolecular hydrogen bonds and hydrophobic interactions, and forms fibrous structures called microfibrils.1,2 The width of the microfibrils depends on the biologically intrinsic arrangements of TCs and is 3-4 nm for plants and 10-20 nm for animals and algae.1,2 Thus, cellulose microfibrils can be defined as advanced bionanofibers. Recent advances in nanotechnology enable materials and devices to be fabricated at the nanoscale level. One of the motivations for the miniaturization process is the superior mechanical properties that nanosized materials possess compared with bulk materials. Nanofibers, in particular, have been used for a wide range of applications such as tissue engineering, filter media, and reinforcement. Such fibers can be made from a variety of materials including synthetic polymers, carbon, and semiconductors in the form of continuous nanofibers, nanofibrous networks, or short nanowires, nanowhiskers, and nanotubes.3 In the case of polymers, the electrospinning process is well-known for fabricating nanofibers. Electrospinning has been applied to many polymers and gives fibers with submicrometer diameters.4 However, very thin nanofibers with a width less than 100 nm, comparable to cellulose microfibrils, are difficult to produce by electrospinning, even if cellulose molecules are used as the nanofiber sources.5 Consequently, understanding all aspects of the characteristics of cellulose microfibrils is quite important for our future. Cellulose microfibrils can be separated from various biological sources by mechanical and chemical treatments. Mechanical treatments that have been used include high-pressure homogenization,6,7 grinding,8-10 and microfluidization.11 These are useful and efficient processes because no additives are required. * To whom all correspondence should be addressed. Tel.: +81-3-58417888. Fax: +81-3-5841-1304. E-mail: [email protected].

However, mostly nanoscaled bundles of microfibrils are obtained, rather than individual microfibrils. On the other hand, chemical treatments can easily disintegrate cellulose into individual microfibrils via modification of the surface properties. Hydrolysis by sulfuric acid is the most common approach, giving short microfibrils called cellulose whiskers.12 A recent advance in chemical treatment was reported by Saito et al.13,14 who achieved isolation of long cellulose microfibrils in high yields through TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl radical)-mediated oxidation. The microfibrils produced by TEMPO-oxidation formed transparent water suspensions and cast-films.15 Due to the very small diameter of the cellulose microfibrils, direct measurement of their mechanical properties has not been reported. Thus, the ideal properties of native cellulose crystals were often referred to as the mechanical properties of the microfibrils. The ability of an atomic force microscopy (AFM) system to accurately apply force in the nano- and pico-Newton range and to measure deformation in the range of Angstroms makes AFM a very useful tool for the mechanical testing of nanosize materials such as cellulose microfibrils. The working principle of AFM is based on the interaction between an AFM cantilever and a sample in contact or near-contact. Utilizing AFM to obtain the elastic modulus of nanoscale fibers by the three-point bending test is a relatively new method. A cantilever is used to apply a small load at the center of a nanofiber suspended over a small groove. The method has been used to obtain an elastic moduli of nanosize beams such as carbon nanotubes,16 carbon nanofibers,17 collagen fibrils,18 β-chitin fibers,19 poly(L-lactic acid) nanofibers,3 bacterial cellulose fibers,20 and regenerated cellulose nanofibers.21 The objective of this study was to measure the elastic modulus of single cellulose microfibrils by a three-point bending test using AFM. Tunicate, which is a marine animal, was used as the source of microfibrils. Due to its high crystallinity, tunicate cellulose has also been used as a model structure to determine crystal and molecular structure.22 The tunicate microfibrils were prepared by two chemical methods, namely, TEMPO-mediated oxidation and sulfuric acid hydrolysis. The microfibrils were

10.1021/bm900520n CCC: $40.75  2009 American Chemical Society Published on Web 07/31/2009

2572

Biomacromolecules, Vol. 10, No. 9, 2009

Iwamoto et al.

deposited on specially designed substrates with nanoscale grooves and subjected to the three-point bending test.

Experimental Section Preparation of Tunicate Microfibrils. Tunicate (Halocynthia papillosa), purchased from a fish store in Tokyo, Japan, was used as starting material. Mantles were separated from the tunicate and cut into small pieces. The cellulose was purified using sodium chlorite and sodium hydroxide, according to the delignification method reported by Yano et al.23 The purified cellulose was disintegrated into microfibrils by both TEMPO-mediated oxidation and sulfuric acid hydrolysis and subsequent mechanical treatment in water. TEMPO-mediated oxidation of cellulose was carried out according to the method reported by Saito et al.13 Cellulose (1 g) was suspended in water (100 mL) containing TEMPO (0.1 mmol) and sodium bromide (1 mmol). The reaction was initiated by adding sodium hypochlorite (5 mmol) at room temperature. The solution was kept at pH 10 by titration with 0.5 M sodium hydroxide for 1 h. At the end of the reaction, the oxidized cellulose was washed with distilled water by filtration. Sulfuric acid hydrolysis of cellulose was performed by treatment with 55 wt % H2SO4 for 2 h at 50 °C under intense stirring, according to the method described elsewhere.24 After hydrolysis, the sample was washed with distilled water by repeated centrifugation at 9000 g for 5 min until it became neutral. The TEMPO-oxidized and acid-hydrolyzed celluloses were fibrillated into individual microfibrils in excess water by homogenizing and sonification for a few minutes. Unfibrillated cellulose fractions were removed from the suspension by centrifugation at 9000 g for 5 min. Finally, translucent suspensions of the microfibrils prepared by TEMPOoxidation and acid hydrolysis were obtained. Atomic Force Microscopy. The atomic force microscope (AFM) used was a scanning probe system comprising SPA-300 and SN-3800 (SII nanotechnology Inc., Japan) units and a cantilever (SI-AF-01, SII nanotechnology Inc., Japan). A drop of the microfibril suspension (0.002 wt %) was put on fresh flat surfaces of mica substrates and the excess was absorbed with blotting paper. After drying at room temperature, the microfibrils on the mica substrates were observed using contact mode AFM. The thicknesses were determined for 20 microfibrils, except for small-sized fractions and domains of kinks or twists. A three-point bending test was performed by AFM to determine the elastic modulus of single microfibrils. Microfabricated substrates with nanoscale grooves were prepared by a photolithographic technique at VDEC, The University of Tokyo. Silicon wafers were coated with photoresists (OAP and ZEP-520A, products of OHKA KOGYO CO. Ltd., Japan and ZEON CORPORATION, Japan, respectively) by spin coating and dried at 180 °C for 15 min. The coated Si wafers were exposed to an electron beam through 120 nm-wide line patterns. The exposed photoresists were removed by soaking the coated Si wafer in a ZED-N50 (ZEON CORPORATION, Japan) solution for 90 s and subsequently in a ZMD-B (ZEON CORPORATION, Japan) solution for 30 s. The line pattern area of the Si wafer was etched with CF4 and C5F8 plasmas for 60 s, and the residual photoresists were removed with oxygen plasma for 60 s. The groove-etched Si wafers were utilized as substrates for the three-point bending test. Scanning electron micrographs of the substrate are shown in Figure 1. The straight line grooves had uniform width 227 nm and depth 1.5 µm determined from scanning electron microscopy (SEM) images. To position the microfibril beams over the grooves, a drop of microfibril suspension (0.002 wt %) was put on the substrates. The excess was absorbed with blotting paper, and the samples were dried at room temperature. The substrates were treated with a glow-discharge for a minute before the dropping of the suspension. The glowdischarging treatment contributed an increase of hydrophilicity to the surface of the substrates, indicating an enhancement of adhesion between the microfibrils and the substrates. Figure 2 shows an SEM image of microfibrils on the substrate. Force curves, which are plots of cantilever deflection with respect to z-piezo displacement, were measured at the middle positions of the microfibril beams and Si wafer

Figure 1. Scanning electron micrographs of the (a) surface and the (b) cross-section of a microfabricated silicon wafer.

Figure 2. Atomic force micrographs of the microfibrils prepared by TEMPO-oxidation of tunicate cellulose on the microfibricated silicon wafer.

after small area observation of the samples. Because the center of the microfibril beam was difficult to determine precisely from the images, the position giving the largest deflection was defined as the center position among three positions close to the center of the beam at 10 nm separation from each other. The force curves were measured for 16 microfibrils, except for domains of kinks or twist using a cantilever (SI-AF-08; SII nanotechnology Inc., Japan), with a spring constant 0.67

Elastic Modulus of Single Cellulose Microfibrils

Figure 3. Transmission electron micrographs of the microfibrils prepared by (a) TEMPO-oxidation and (b) acid hydrolysis of tunicate cellulose.

N m-1 measured by the thermal noise technique25 using an AFM (MFP3D-SA, Asylum Technology Co., Ltd.). The same cantilever was used for the all-force curve measurements. Transmission Electron Microscopy. A 10 µL aliquot of 0.02 wt % microfibril suspension was mounted on a glow-discharge carboncoated electron microscopy grid. The excess liquid was absorbed by filter paper, and a drop of 2% uranyl acetate negative stain was added before drying. Excess solution was blotted with a filter paper and allowed to dry by evaporation under ambient conditions. The sample grid was observed at 100 kV using a JEOL electron microscope (JEM 2000-EXII). Average widths of 20 microfibrils, except for small-size fractions and domains of kinks or twists, were determined from the transmission electron microscopy (TEM) images. X-ray Diffraction. X-ray diffraction experiments were carried out using beamline BL45XU with wavelength 0.09 nm at the SPring-8 synchrotron radiation facility. The microfibril suspension (0.1 wt %) was dried at 105 °C on Petri dishes of PTFE to obtain films of ca. 100 µm thickness. Two dimensional wide-angle X-ray diffraction patterns were recorded using a charge coupled device (CCD) camera with exposure time 73 ms. The pixel size of the CCD camera was 125 × 125 µm, with 12 bits per pixel. The camera length for measurements was 140 mm. The two-dimensional patterns obtained were converted into one-dimensional patterns by image analysis.

Results and Discussion Morphology and Crystallinity of Tunicate Microfibrils. Figure 3 shows TEM micrographs of the microfibrils prepared by (a) TEMPO-oxidation and (b) acid hydrolysis. They were

Biomacromolecules, Vol. 10, No. 9, 2009

2573

well separated and had some domains of twists and kinks. It seemed that the twists reflected the native structure of tunicate microfibrils, and the kinks were probably due to damage by the mechanical treatment. The oxidation of cellulose using TEMPO catalyst in aqueous media changes the C6 primary hydroxyl groups to carboxylate groups. Regenerated cellulose has been completely converted to water-soluble polyglucuronic acid by oxidation.26,27 In the case of native celluloses, oxidation occurs only on the surface of microfibrils, and the oxidized microfibrils retain the native crystalline structure.13 Furthermore, because the cellulose microfibrils produced by TEMPO-oxidation have surface carboxylate groups, there are repulsive forces between the microfibrils in aqueous media. Consequently, the individual microfibrils are easily dispersed in water by simple mechanical treatment such as stirring or sonification. Saito et al.13 reported procedures for preparation of microfibrils by TEMPO-oxidation of cellulose from a variety of native sources including pulp, cotton, tunicate, and bacterial cellulose. In the present study, the microfibrils (Figure 3a) from TEMPOoxidation were 20.3 (standard deviation, σ ) 1.6) nm wide and 8.4 (σ ) 0.6) nm thick, as determined from TEM and AFM images, respectively (Table 1). The length was difficult to measure, but the longest microfibril had length of about 10 µm from AFM images. Acid hydrolysis with sulfuric acid selectively degrades amorphous regions of celluloses, and small amounts of sulfate ester groups are introduced on the surfaces of cellulose crystals. Rod-like crystals (whiskers) are obtained in aqueous media by subsequent mechanical treatment. In this study, cellulose nanofibers with a much larger aspect ratio than the whiskers were obtained by acid hydrolysis (Figure 3b) because the treatment time was shorter than in the reported method.24 This indicated that the amorphous regions of microfibrils were not completely removed. These microfibrils were 19.9 (σ ) 1.6) nm wide and 8.8 (σ ) 0.6) nm thick from TEM and AFM images, respectively (Table 1). The longest microfibril was about 5 µm long, shorter than that of the longest microfibril prepared by TEMPO-oxidation, due to the degradation of amorphous regions of tunicate cellulose by acid hydrolysis. Figure 4 shows X-ray diffraction patterns of the microfibrils that were prepared by TEMPO-oxidation and acid hydrolysis of tunicate cellulose. Only haloes from amorphous regions were found in the diffraction patterns of both types of microfibrils, indicating that their crystallinities were extremely high. Five peaks were found in the diffraction patterns, and were identified as from (1 1j 0), (1 1 0), (1 1 1), (1 0 2) + (0 1 2), and (2 0 0) planes.28 The diffraction patterns were typical of those of native cellulose crystals. Three-Point Bending Test of the Microfibrils. The microfibril suspension was dried on the microfabricated silicon wafer substrates and deposited over the nanoscale etched grooves, as shown in Figure 3. The microfibrils on the substrates kept the same position during the observations and the measurements. Therefore, the microfibrils were assumed to be enough to fix on the substrates for the mechanical tests. The cantilever deflections were measured on the microfibrils in the middle position of the span and the substrates to obtain the force curves (Figure 5), which represent the deflection behavior of the cantilever against the z-piezo displacement. The curves in Figure 5 were extracted without noncontact region from a whole curve, and the origin of the force curve was set at the contact point of the cantilever with the microfibril and the substrate. The range of the force curve measurements was in a linear region of the curves. It was confirmed that the elastic limit of the microfibrils

2574

Biomacromolecules, Vol. 10, No. 9, 2009

Iwamoto et al.

Table 1. Characteristics of the Microfibrils Prepared by (a) TEMPO-Oxidation and (b) Acid Hydrolysis of Tunicate Cellulose sample

width (nm)

(a) TEMPO-oxidation (b) acid hydrolysis ramie ramie ramie flax

20.3 (1.6 ) 19.9 (1.6a) microfibers microfibers microfibers microfibers

Iβ Iβ Iβ tunicate

whiskers

a

thickness (nm)

a

The standard deviations of 20 samples.

a

8.4 (0.6 ) 8.8 (0.6a)

b

length (µm)

technique

∼ca. 10 ∼ca. 5

AFM AFM X-ray33 X-ray34 X-ray35 X-ray36 theoretical37 theoretical38 theoretical39 theoretical40 theoretical41 raman spectroscopy42

elastic modulus (GPa) 145.2 (31.3b) 150.4 (28.8b) 134 120-135 138 220 ( 50 136 ( 6 167.5 149, 116 124-155 156 143

The standard deviations of 16 samples.

Figure 4. X-ray diffraction patterns (λ ) 0.09 nm) of the microfibrils prepared by (a) TEMPO-oxidation and (b) acid hydrolysis of tunicate cellulose. Arrows indicate identified peaks.

Figure 6. Cross-sectional shape models of the microfibrils prepared by (a) TEMPO-oxidation and (b) acid hydrolysis proposed by Helbert et al.29 The ellipse in (b) indicates an approximation of the eroded cross-sectional shape.

the microfibril was calculated from the product of the deflection of the cantilever on the microfibril (β in Figure 5) and the spring constant of the cantilever (0.67 N m-1). The elastic modulus (E) was calculated based on the elastic beam theory for three-point bending of a beam with two ends fixed and is given by Figure 5. Force curves on the silicon wafer (solid line) and a microfibril over the groove (dotted line). R indicates the deflection of the microfibril and β is used for the calculation of the loaded force.

E ) F · L3 /192 · D · I

was not exceeded during the tests. The z-piezo displacement beyond the contact point of cantilever was typically about 60 nm. The deformation of the substrate occurring during force curve measurement could be eliminated, because the Si wafer was sufficiently stiff for its deformation to be negligible. Accordingly, the deflection of the cantilever on the substrate was assumed to be equal to the z-piezo displacement. On the other hand, because the microfibril was bent by the cantilever, the deflection of the cantilever on the microfibril was attributed to the difference between the z-piezo displacement and the deflection of the microfibril. Hence, the deflection of the microfibril (D) was determined by subtracting the deflection of the cantilever on the microfibril from that on the substrate, as denoted by R in Figure 5. In addition, the loaded force (F) on

where L and I are the span length and the second moment of area of the microfibril, respectively. The span lengths determined as 227-264 nm, by calculations from the angle between the microfibril and the groove. The second moment of area depends on the cross-sectional shape of the beam. Helbert et al.29 reported that native tunicate microfibrils have a parallelogram cross-sectional shape, as shown in Figure 6a. It was assumed that the microfibrils prepared by TEMPO-oxidation maintained the native crosssectional shapes because TEMPO-mediated oxidation was carried out under mild conditions at pH 10 and room temperature. Thus, the second moment of area of those microfibrils (I1) was calculated from the equation

Elastic Modulus of Single Cellulose Microfibrils

Biomacromolecules, Vol. 10, No. 9, 2009

2575

I1 ) b1 · h31 /12 ) 712.1 nm4 where b1 and h1 represent the width and height of the parallelogram, respectively. On the other hand, it was reported by Helbert et al.29 that acid hydrolysis eroded edges of the microfibrils, as depicted in Figure 6b. On that basis, the cross-sectional shapes of the microfibrils prepared by acid hydrolysis were assumed to be approximately elliptical. Thus, the second moment of area of these microfibrils (I2) was obtained using the equation

I2 ) π · b1 · h32 /64 ) 665.7 nm4 where b2 and h2 represent the major and minor diameters of the ellipse. The average values of the width from TEM observations and thickness from AFM images of the microfibrils from TEMPO-oxidation and acid hydrolysis were used to determine their second moment of area. The reason for adoption of that procedure was that accurate widths could not be obtained from AFM images at the force curve measurement positions, due to unknown conditions of cantilever probes. Figure 7 shows the elastic moduli of the microfibrils prepared by the two methods. The average values ((σ) of 16 samples were 145.2 ( 31.3 and 150.7 ( 28.8 GPa for the microfibrils prepared by TEMPO-oxidation and acid hydrolysis, respectively (Table 1). The difference in the average moduli of the microfibrils prepared by the two methods was not significant. The standard deviations (σ) were probably due to the uncertainties in the cross-sectional areas, which were not determined for the microfibrils on the grooved substrate. Because the thickness, in particular, appears as a cubic term in the second moment of area, this has the potential of introducing a fairly large uncertainty. In general, deflections of a beam involve both bending and shear deformations. The ratio bending/shear deformation depends on the ratio of the beam length to its diameter. In this study, the beam length varied from 227 to 264 nm because of the angles between the microfibrils and grooves. The beam diameter was estimated as 14 nm (the mean of 20 nm width and 8 nm thickness), hence the beam length/diameter ratio was in range 16-19. However, the elastic moduli did not depend on the length/diameter ratio, indicating that the shear deformation could be eliminated from the deflection of the microfibrils. The reason is probably the uniform molecular structure of the microfibril, i.e. aggregations of extended cellulose molecular chains. Elastic moduli of native cellulose crystals, determined using several techniques, have been previously reported (Table 1). The early studies were achieved with theoretical30,31 and X-ray analysis.32,33 It is well-known that X-ray analysis by Sakurada et al.33 reported 134 GPa of the crystal modulus. Since then, the following studies showed that the values based on X-ray analysis were 120-135,34 138,35 and 220 ( 50 GPa36 and theoretical calculations gave 136 ( 6,37 167.5,38 149 or 116,39 124-155,40 and 156 GPa.41 Sturcova et al.42 recently used Raman spectroscopy and obtained 143 GPa for the elastic modulus of tunicate cellulose whiskers. The elastic moduli of tunicate microfibrils measured by AFM bending tests were similar to the previously reported values. The crystallinities of the tunicate microfibrils were extremely high, as shown in the X-ray diffraction patterns (Figure 4). In addition, the AFM bending experiments were conducted on the straight portions of microfibrils. The domains of the twists or kinks were presumed to include disordered regions of cellulose molecular

Figure 7. Elastic moduli of the microfibrils prepared by (a) TEMPOoxidation and (b) acid hydrolysis of tunicate cellulose.

orientation. Therefore, it was concluded that, for the tunicate microfibrils, the bending modulus using AFM for the domains consisting mainly of crystals demonstrated good agreement with the modulus from the previous reports determined using other techniques.

Conclusion This study gave an experimentally determined elastic modulus for single microfibrils of cellulose, by the threepoint bending test using AFM. Tunicate, which is a marine animal, was used as source of cellulose microfibrils. The tunicate microfibrils were prepared by two chemical methods, namely, TEMPO-mediated oxidation and sulfuric acid hydrolysis. The microfibrils had cross-sectional dimensions of the order 8 × 20 nm and lengths several micrometers. They were deposited on specially designed substrates with nanoscaled grooves for the three-point bending tests using an AFM cantilever. The elastic moduli of the single microfibrils prepared by TEMPO-oxidation and acid hydrolysis were 145.2 and 150.7 GPa, respectively. The bending tests using AFM for the tunicate microfibrils gave elastic moduli similar to previously reported values for native cellulose crystals. Acknowledgment. We would like to thank Assistant Prof. Tsuguyuki Saito, Department of Biomaterial Science, Graduate School of Agricultural and Life Sciences, The University of Tokyo, for the TEM observations. We also thank Asylum Technologies Co. Ltd. for the calibrations of the cantilever spring constant. The preparation of microfabricated substrates

2576

Biomacromolecules, Vol. 10, No. 9, 2009

was conducted in Center for Nano Lithography and Analysis, The University of Tokyo, supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. Finally, this study was supported by the New Energy and Industrial Technology Development Organization (NEDO), Japan.

References and Notes (1) Osullivan, A. C. Cellulose 1997, 4, 173–207. (2) Brown, R. M. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 487– 195. (3) Tan, E. P. S.; Lim, C. T. Nanotechnology 2006, 17, 2649–2654. (4) Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S. Compos. Sci. Technol. 2003, 63, 2223–2253. (5) Kim, C. W.; Kim, D. S.; Kang, S. Y.; Marquez, M.; Joo, Y. L. Polymer 2006, 47, 5097–5107. (6) Turbak, A. F.; Snyder, F. W.; Sandberg, K. R. J. Appl. Polym. Sci.: Appl. Polym. Symp. 1983, 37, 815–827. (7) Herrick, F. W.; Casebier, R. L.; Hamilton, J. K.; Sandberg, K. R. J. Appl. Polym. Sci.: Appl. Polym. Symp. 1983, 37, 797–813. (8) Taniguchi, T.; Okamura, K. Polym. Int. 1998, 47, 291–294. (9) Iwamoto, S.; Nakagaito, A. N.; Yano, H. Appl. Phys. A: Mater. Sci. Process. 2007, 89, 461–466. (10) Abe, K.; Iwamoto, S.; Yano, H. Biomacromolecules 2007, 8, 3276– 3278. (11) Zimmermann, T.; Pohler, E.; Geiger, T. AdV. Eng. Mater. 2004, 6, 754–761. (12) Samir, M.; Alloin, F.; Dufresne, A. Biomacromolecules 2005, 6, 612– 626. (13) Saito, T.; Nishiyama, Y.; Putaux, J. L.; Vignon, M.; Isogai, A. Biomacromolecules 2006, 7, 1687–1691. (14) Saito, T.; Kimura, S.; Nishiyama, Y.; Isogai, A. Biomacromolecules 2007, 8, 2485–2491. (15) Fukuzumi, H.; Saito, T.; Wata, T.; Kumamoto, Y.; Isogai, A. Biomacromolecules 2009, 10, 162–165. (16) Salvetat, J. P.; Briggs, G. A. D.; Bonard, J. M.; Bacsa, R. R.; Kulik, A. J.; Stockli, T.; Burnham, N. A.; Forro, L. Phys. ReV. Lett. 1999, 82, 944–947. (17) Lawrence, J. G.; Berhan, L. M.; Nadarajah, A. ACS Nano 2008, 2, 1230–1236.

Iwamoto et al. (18) Yang, L.; van der Werf, K. O.; Koopman, B.; Subramaniam, V.; Bennink, M. L.; Dijkstra, P. J.; Feijen, J. J. Biomed. Mater. Res. 2007, 82A, 160–168. (19) Xu, W.; Mulhern, P. J.; Blackford, B. L.; Jericho, M. H.; Templeton, I. Scanning Microsc. 1994, 8, 499–506. (20) Guhados, G.; Wan, W. K.; Hutter, J. L. Langmuir 2005, 21, 6642– 6646. (21) Cheng, Q. Z.; Wang, S. Q. Composites, Part A 2008, 39, 1838–1843. (22) Nishiyama, Y.; Langan, P.; Chanzy, H. J. Am. Chem. Soc. 2002, 124, 9074–9082. (23) Yano, H.; Hirose, A.; Collins, P. J.; Yazaki, Y. J. Mater. Sci. Lett. 2001, 20, 1125–1126. (24) Elazzouzi-Hafraoui, S.; Nishiyama, Y.; Putaux, J. L.; Heux, L.; Dubreuil, F.; Rochas, C. Biomacromolecules 2008, 9, 57–65. (25) Hutter, J. L.; Bechhoefer, J. ReV. Sci. Instrum. 1993, 64, 1868–1873. (26) Isogai, A.; Kato, Y. Cellulose 1998, 5, 153–164. (27) Tahiri, C.; Vignon, M. R. Cellulose 2000, 7, 177–188. (28) Sugiyama, J.; Vuong, R.; Chanzy, H. Macromolecules 1991, 24, 4168– 4175. (29) Helbert, W.; Nishiyama, Y.; Okano, T.; Sugiyama, J. J. Struct. Biol. 1998, 124, 42–50. (30) Meyer, K. H.; Lotmar, W. HelV. Chim. Acta 1936, 19, 68–86. (31) Treloar, L. R. G. Polymer 1960, 1, 290–303. (32) Mann, J.; Roldangonzalez, L. Polymer 1962, 3, 549–553. (33) Sakurada, I.; Nukushina, Y.; Ito, T. J. Polym. Sci. 1962, 57, 651– 660. (34) Matsuo, M.; Sawatari, C.; Iwai, Y.; Ozaki, F. Macromolecules 1990, 23, 3266–3275. (35) Nishino, T.; Takano, K.; Nakamae, K. J. Polym. Sci., Part B: Polym. Phys. 1995, 33, 1647–1651. (36) Diddens, I.; Murphy, B.; Krisch, M.; Muller, M. Macromolecules 2008, 41, 9755–9759. (37) Kroon-batenburg, L. M. J.; Kroon, J.; Northolt, M. G. Polym. Commun. 1986, 27, 290–292. (38) Tashiro, K.; Kobayashi, M. Polymer 1991, 32, 1516–1530. (39) Eichhorn, S. J.; Davies, G. R. Cellulose 2006, 13, 291–307. (40) Tanaka, F.; Iwata, T. Cellulose 2006, 13, 509–517. (41) Bergenstrahle, M.; Berglund, L. A.; Mazeau, K. J. Phys. Chem. B 2007, 111, 9138–9145. (42) Sturcova, A.; Davies, G. R.; Eichhorn, S. J. Biomacromolecules 2005, 6, 1055–1061.

BM900520N