Biomacromolecules 2010, 11, 1281–1285
1281
Converse Piezoelectric Effect in Cellulose I Revealed by Wide-Angle X-ray Diffraction Wolfgang Gindl,*,† Gerhard Emsenhuber,† Johannes Plackner,† Johannes Konnerth,† and Jozef Keckes‡ Department of Material Sciences and Process Engineering, BOKU-University of Natural Resources and Applied Life Science, A-1190 Vienna, Austria, and Department of Materials Physics, University of Leoben and Erich Schmid Institute for Materials Science, Austrian Academy of Sciences, A-8700 Leoben, Leoben, Austria Received January 21, 2010; Revised Manuscript Received March 19, 2010
The converse piezoelectric effect in cellulose I was studied by exposing thin pine wood slices to an electric field. Macroscopically, a strong extension of wood was observed in its transverse anatomical direction (grain angle 90°), perpendicular to the direction of the electric field. The same effect, albeit to a lesser extent, was observed for specimens with a 45° grain angle, whereas no measurable dimensional change was observed for specimens with grain oriented parallel to the testing direction (0° grain angle). The measured extension in the transverse direction was proportional to the intensity of the applied electric field and amounted to 0.0278% on average at a field intensity of 1 MV m-1, which results in a piezoelectric charge constant of 278 pm V-1. At the nanoscale, changes in the cellulose crystallites due to the applied electric field were studied by means of wide-angle X-ray diffraction using the same specimens as in macroscopic experiments. Significant radial shifts of the scattering intensity peak attributed to the cellulose 200 crystallographic plane toward smaller scattering angles were observed, while the electric field was applied. These peak shifts were attributed to an increase in the spacing of the 200 crystallographic planes of cellulose I. At an electric field intensity of 1 MV m-1, a crystallite strain ε⊥200 normal to the 200 reflection plane of 0.2% was estimated from Bragg’s law.
Introduction When a certain type of crystal such as quartz is exposed to mechanical stress, an electrical polarization is observed, which is termed direct piezoelectric effect. When, on the other hand, an electrical field is applied to a crystal, mechanical strain results. The latter phenomenon is called converse piezoelectric effect.1 Besides well-known and widely applied piezo-electric materials such as quartz or tourmaline, also biogenic materials like collagen, bone, and wood are known to exhibit piezoelectricity.2-5 Fukada1 and Bazhenov4 were the first to deliver a comprehensive description of mostly direct macroscopic piezoelectric phenomena in wood. The most important results of their experiments were that wood exhibits both direct and converse piezoelectricity, that the piezoelectric effect is strongest when wood is loaded at an angle of 45° to the direction of the wood fibres, and that a reversal of the load direction results in a change of electric field polarity. Following the work of Fukada and Bazhenov, studies dedicated to wood piezoelectricity were mainly focused on potential technological applications of this effect, mainly in nondestructive testing.6-9 Recently, interest in piezoelectricity of cellulose was renewed by a series of publications on electro-active paper.10-15 In good agreement with results on wood, it was found that electroactive paper shows both direct and converse piezoelectricity, which is strongest at an angle of 45° with regard to the direction of preferred orientation of cellulose chains and is polaritydependent just as in wood. This good agreement is remarkable insofar as the constituent material of electro-active paper is * To whom correspondence should be addressed. Tel.: ++43-1-476544255. Fax: ++43-1-47654-4295. E-mail:
[email protected]. † BOKU-University of Natural Resources and Applied Life Science. ‡ Austrian Academy of Sciences.
regenerated cellulose (cellulose II), which is different from the cellulose I crystal conformation that is characteristic of wood. The intriguing results published by Kim and his co-workers cited above and aptly summarized by the title of their 2006 paper “Discovery of cellulose as a smart material”10 make it seem opposite to revisit piezoelectricity in wood, to evaluate whether this natural material is capable of delivering similar effects usable for biobased actuators as described for electroactive paper. Because converse piezoelectricity of wood would be most interesting with regard to actuators, it is the scope of the study presented in this paper.
Experimental Section Slices of pine wood with a thickness of 0.2 mm, a length of 100 mm, and a width of 8 mm were prepared by means of a microtome. The slices were cut along the radial anatomical plane in a way that the wood fibres were oriented at 0, 45, and 90° with respect to the specimen length (Figure 1). The slices were conditioned to a moisture content of 12% and 10 specimens each were set aside for tensile testing on a Zwick 20 kN universal testing machine. At a cross head displacement of 1 mm min-1, the specimen elongation was recorded by means of a Zwick macrosense clip-on displacement sensor for the calculation of the modulus of elasticity. For experiments with a DC electric field, the radial surfaces of the wood slices were covered with conductive silver paint (Silber Leitlack 5900, Busch GmbH, Viernheim, Germany). The use of silver paint required a minimum specimen thickness of 0.2 mm because electric short-cuts occurred frequently with thinner specimens. Even though short-cuts occasionally occur at this specimen thickness, a further increase in thickness was not feasible, because already at 0.4 mm, no piezoelectric effect could be observed. Thereafter, the slices were fixed to rubber-coated and, thus, electrically insulating grips of the universal testing machine. Prior to applying an electric field by means of
10.1021/bm1000668 2010 American Chemical Society Published on Web 03/30/2010
1282
Biomacromolecules, Vol. 11, No. 5, 2010
Gindl et al.
Figure 1. Set-up for the measurement of the macroscopic converse piezoelectric effect. Wood specimens of different fiber angle coated with silver lacquer were fixed to the grips of a tensile testing machine. While the machine displacement (d) was held constant, variations in the force (F) due to changes in the voltage (U) over time (t) were recorded by a load cell.
The obtained 2D detector images were spatially corrected and subsequently analyzed by means of Fit2D (www.esrf.eu/computing/ scientific/FIT2D/) software.
Results and Discussion
Figure 2. Plot of tensile stress in a wood specimen during a piezoelectric experiment. After loading the specimen to a target stress of 2 MPa the machine cross head is held at constant displacement while continuously recording stress. Due to the application of an electrical field, the stress recorded by the tensile testing machine changes (inset a). In parallel to changes due to piezoelectricity, the recorded stress diminishes continuously due to stress relaxation. This effect is removed mathematically by detrending (inset b).
contacting the silver coating on both radial wood surfaces, the sample was strained to 0.2%, and the cross-head position was held constant during the following experiment. A 50 N load cell was used to record eventual load changes due to changes in specimen length induced by the applied electric field (Figure 1). As shown in Figure 2, pronounced stress relaxation occurred during such experiments. This long-term change in recorded stress was removed from each data set by means of fitting an exponential decay function of the shape
σ(t) ) a + e-bt
(1)
where σ is the tensile stress, t is time, and a and b are the fitted parameters. After subtracting the fitted trend line from the actually measured stress graph, residual variations in stress were evaluated with regard to potential piezoelectric effects. For control, a polyethylene film with a thickness of 0.1 mm was subjected to the same procedure as wood specimens. The same specimens used for macroscopic experiments were also used for wide-angle X-ray diffraction on a Bruker AXS Nanostar. Silver-coated wood slices were positioned in the vacuum chamber of the Nanostar in a way that the X-ray beam passed the sample perpendicularly to the silver-coated radial plane. Measurements were taken both with and without applying an electric field to the specimen.
Converse Piezoelectricity of Thin Wood Slices. Depending on the angle between the direction in which changes in mechanical stress were recorded and the direction of wood grain, different results shown in Figure 3 were obtained upon application of an electric field. To draw Figure 3, the measured stress was converted to strain assuming linear elasticity using the experimentally determined values of 10, 1.2, and 0.5 GPa for the modulus of elasticity at 0, 45, and 90° grain angles, respectively. At an angle of 0° between the direction of mechanical stress measurement and wood grain, no measurable effect was discerned upon switching on and off the electric field. Rotation of the grain to an angle of 45° resulted in a significant elongation of the specimen when the electric field was switched on. For about 10 s, the specimen continued to elongate until saturation was achieved. Switching off the electric field leads to the full recovery of the original specimen length. A clearly stronger effect of the electric field was observed at a grain angle of 90°, where an elongation to 0.0278% strain occurred at a field intensity of 1 MV m-1. Similar experiments conducted with reversed polarity of the applied electric field yielded the same results, as described above, that is, no effect of reversed polarity was observed. Control experiments performed with a silver-coated polyethylene film did not result in any measurable mechanical signal. This observation is interpreted as confirming that piezoelectric effects measured in the present study originate from wood and not from the silver coating applied to the wood surface. To determine a piezoelectric charge constant for the specific setup used in the present study, a set of experiments was conducted at 90° grain orientation, applying different electric field intensities. The results of this experiment shown in Figure 4 indicate a linear relationship between the intensity of the applied electric field and the resulting specimen strain. From the slope of a linear regression function fitted to the data shown on Figure 4, a piezoelectric charge constant of 278 pm V-1 was calculated. The results shown in Figure 3 are in conflict with earlier literature on wood piezoelectricity insofar as a strong effect was measured not only at 45° grain angle, which is in good
Converse Piezoelectric Effect in Cellulose I
Figure 3. Electric field-induced strain for wood specimens at 0, 45, and 90° fiber angle. Filled and empty triangles, respectively, indicate switching on and off the electric field. Strain was calculated from the recorded load on the load cell and the respective elastic modulus assuming linear elastic behavior.
Figure 4. Electric field-induced strain calculated for a wood specimen at 90° fiber angle exposed to different field intensities. Strain was calculated from the recorded load on the load cell and the respective elastic modulus assuming linear elastic behavior.
agreement, but also at 90°, where no effect should be observed according to Bazhenov4 and Fukada.2 Also, no measurable effect of reversed polarity, as claimed by Bazhenov4 and Fukada,2 was observed in the present study. We have no explanation for this discrepancy, but it should be noted that a direct comparison of the results shown in Figure 3 with existing literature on wood piezoelectricity is probably not possible because past experiments mostly refer to direct piezoelectricity observed with specimens several centimeters thick,4 whereas converse piezoelectricity was studied here, using thin and well-defined microsections. It is assumed that the porosity and complex anatomical structure of wood counteract the build-up of a homogeneous electric field throughout specimens of any size,
Biomacromolecules, Vol. 11, No. 5, 2010
1283
which hinders comparability of experiments performed with specimens of significantly different size. Converse Piezoelectricity of Cellulose I. Wide angle X-ray diffraction 2D detector images, as shown in Figure 5, were evaluated to identify potential differences in the cellulose I crystalline structure caused by an applied electric field. Proceeding from small to large diffraction angles, the detector images allowed to determine the radial and the azimuthal position, respectively, of the overlapping 110/110 reflection and the 200 reflection, which all are observed normal to the c-crystallographic axis of cellulose, and the 004 reflection parallel to the c-axis (Figure 6). By comparing detector images obtained with and without an applied electric field, no measurable differences were found in the azimuthal position of the diffraction peaks mentioned above. The absence of azimuthal peak shifts upon application of an electric field speaks against electrically induced shear deformation of the cellulose crystallite. Also, no change in azimuthal peak width indicative of changes in cellulose fibril angle was found. In contrast to azimuthal peak positions, small but clear radial shifts of diffraction peaks were found. Using the distinct silver 111 reflection for reference, a shift of -0.06° caused by an applied electric field with an intensity of 1 MV m-1 was observed for the cellulose 200 reflection. A significant but less clear shift of smaller magnitude in the same direction was found for the overlapping 110/110 reflection and no measurable shift was seen for the 004 reflection. From an evaluation of the shift in the 200 scattering peak position by means of peak fitting performed with three independent specimens, the concurrent change in the cellulose I crystallite was estimated using Bragg’s law:2
nλ ) 2d sin θ
(2)
where n is an integer determined by the order given, λ is the wavelength of the X-rays, d is the spacing between the planes in the atomic lattice, and θ is the angle between the incident ray and the scattering planes. This evaluation yielded a crystal strain ε⊥200 normal to the 200 reflection plane of 0.2% at an electric field intensity of 1 MV m-1. This value is of a reasonable order considering that at the macroscopic scale, a strain of 0.028% was measured at the same field intensity. Because wood consists of about 50% cellulose, two-thirds of which is crystalline, and 50% noncrystalline polysaccarides and lignin,16 it is clear that a strong effect has to act at the nanoscale for a moderate effect to manifest itself at the macroscopic scale. In a modeling study on the direct piezoelectric effect in cellulose I, Pizzi and Eaton17 stated that even though a dominant pattern of orientation of cellulose microfibrils is found in the wood cell wall, cellulose crystallites are loaded in many different modes when wood is exposed to mechanical stress. A macroscopically measured direct piezoelectric effect may therefore be the result of a number of superimposed effects caused by different loads on the cellulose crystallites. A similar assumption may be formulated for converse piezoelectricity: the cellulose I crystallite may react with a complex pattern of deformation to an applied electric field. However, from our X-ray diffraction data it is concluded that an extension of the crystallite normal to the 200 crystallographic plane is dominant. This finding is supported by the modeling study cited above,17 which identified primarily nonpolar interactions (van der Waals interactions) being responsible for cellulose piezoelectricity. A scheme of the crystalline unit cell of cellulose I adapted after Bohn18 is shown in Figure 7. In the cellulose I crystal, polar interactions (hydrogen bonds) are only found in the plane spun up by the c
1284
Biomacromolecules, Vol. 11, No. 5, 2010
Gindl et al.
Figure 5. Wide angle X-ray diffraction 2D detector images: left, silver coating; center, pine wood at 45° fiber angle; right, silver-coated pine wood at 45° fiber angle.
Figure 6. Example of an integrated scattering intensity distribution curve for a silver coated wood specimen. The strongest reflections for cellulose Iss and the silver reflection Ag 111 used for reference are indicated. The inset graph shows the small shift in the intensity distribution due to an electric field of 1 MV m-1 for the 110/110 reflection and the 200 reflection.
Figure 8. Softwood in schematic cross-sectional view. An extension of the cellulose-rich secondary cell wall (gray) in transverse direction as symbolized by arrows is proposed. Table 1. Comparison of Selected Properties for Wood and Electroactive Paper property constituents crystalline cellulose degree of orientation (f)
Figure 7. Schematic structure of cellulose Iβ (monoclinic unit cell with γ ) 97.0°, a0 ) 0.817 nm, b0 ) 0.786 nm, c0 ) 1.038 nm, projection in direction of the crystallographic c-axis, which corresponds to the cellulose chain axis).
and a crystallographic axes corresponding to the 200 plane of reflection. By contrast, only nonpolar interactions act along the b-crystallographic axis, which corresponds to the direction of piezoelectric crystal strain normal to the 200 reflection plane found in the present study. In normal wood, cellulose chains are predominantly oriented in the direction of the prosenchymate wood cell axis. The direction of crystallite strain ε⊥200 identified by X-ray diffraction is transverse to the cellulose chain. Therefore, it is highly probable that the crystallite strain ε⊥200 is responsible for the substantial elongation of macroscopic wood specimens observed transverse to the wood fiber axis, as schematically shown in Figure 8.
wood
electroactive paper
cellulose (I), noncellulosic polysaccharides, lignin 30% 0.57
cellulose (II), viscose process 29%
