COMMUNICATION pubs.acs.org/Biomac
Selective Detection of Crystalline Cellulose in Plant Cell Walls with Sum-Frequency-Generation (SFG) Vibration Spectroscopy Anna L. Barnette,† Laura C. Bradley,† Brandon D. Veres,† Edward P. Schreiner,† Yong Bum Park,‡ Junyeong Park,§ Sunkyu Park,§ and Seong H. Kim*,† †
Department of Chemical Engineering and ‡Department of Biology, Pennsylvania State University, University Park, Pennsylvania 16802, United States § Department of Forest Biomaterials, North Carolina State University, Raleigh, North Carolina 27695, United States
bS Supporting Information ABSTRACT: The selective detection of crystalline cellulose in biomass was demonstrated with sum-frequency-generation (SFG) vibration spectroscopy. SFG is a second-order nonlinear optical response from a system where the optical centrosymmetry is broken. In secondary plant cell walls that contain mostly cellulose, hemicellulose, and lignin with varying concentrations, only certain vibration modes in the crystalline cellulose structure can meet the noninversion symmetry requirements. Thus, SFG can be used to detect and analyze crystalline cellulose selectively in lignocellulosic biomass without extraction of noncellulosic species from biomass or deconvolution of amorphous spectra. The selective detection of crystalline cellulose in lignocellulosic biomass is not readily achievable with other techniques such as XRD, solid-state NMR, IR, and Raman analyses. Therefore, the SFG analysis presents a unique opportunity to reveal the cellulose crystalline structure in lignocellulosic biomass.
’ INTRODUCTION Cellulose is the most abundant natural polymer on earth and is an important constituent of pulp, paper, and textile materials. In addition, cellulose can be a renewable resource for biofuels and chemical production. Cellulose in plant cell walls consists of β-Dglucopyranose units polymerized through 1,4-glycosidic linkage. The intrachain and interchain hydrogen bonds hold the chains together edge-to-edge in flat sheets. The sheets are stacked to form crystals with two allomorph structures (IR and Iβ).1 These interactions arrange the chains into fibrils with a diameter of 310 nm in plant cell walls.2 It is very important to study the cellulose structure in its native state for better understanding of plant growth processes and design of more efficient biomass conversion processes. The crystalline structure of cellulose can be characterized using several analytical techniques including X-ray diffraction (XRD), solid-state 13C nuclear magnetic resonance (NMR), infrared (IR), and Raman spectroscopy. XRD is probably the most extensively used technique for cellulose structural study.3,4 It detects diffraction of X-ray from periodic planes with the same electronic densities in the crystalline lattice of cellulose. Amorphous structures cannot diffract X-ray; they produce featureless diffuse scattering. It is particularly noteworthy that the calculation of cellulose crystallinity from XRD requires correction of background contributions from the diffuse scattering from the amorphous carbohydrates.5 The crystallinity calculation from XRD data could give different estimates depending on the background r 2011 American Chemical Society
correction method.3,4 It is shown that the simplest and thus most widely used peak-height method tends to overestimate the true crystallinity.3,4 The accurate calculation of cellulose crystallinity from XRD data is still active research area.3,6 Cross-polarization solid-state 13C NMR can be used to cellulose and its interactions with other carbohydrate polymers in plant cells.7,8 This method utilizes differences in the proton relaxation rates and its coupling with neighboring carbon atoms between crystalline cellulose microfibils and noncellulose polymers. This difference can be utilized to distinguish the signals from crystalline components within rigid frameworks and noncrystalline components with larger mobilities. Although it was demonstrated to work, the method is quite complicated to use for routine analysis. The more widely used method is much simpler 13 C NMR analysis, which measures the chemical shift of the carbon atom involved in the glycosidic bond (C4 position of the glucopyranose ring). However, this widely used NMR routine suffers from the overlap with the amorphous components and requires deconvolution of the crystalline peak from amorphous contribution.3,4 It becomes more challenging to analyze cellulose in biomass because hemicellulose and lignin components can interfere with the cellulose peaks. To avoid this spectral interference, cellulose components can be separated from other Received: April 15, 2011 Revised: May 24, 2011 Published: May 26, 2011 2434
dx.doi.org/10.1021/bm200518n | Biomacromolecules 2011, 12, 2434–2439
Biomacromolecules
COMMUNICATION
amorphous components via extraction and delignification processes; however, these treatments might alter the cellulose crystal structures or their interactions with noncellulose matrix.9 In this Communication, we report direct and selective detection of crystalline cellulose in plant cell walls using visible-IR sum-frequency-generation (SFG) vibration spectroscopy. SFG is a second-order nonlinear optical response from a system where optical centrosymmetry is broken.10 The frequency of SFG signal is the sum of visible and IR frequencies, ωSFG = ωVIS þ ωIR, and its intensity I(ωSFG) is given by eq 111 IðωSFG Þ jχef f ð2Þ j2 IðωVIS ÞIðωIR Þ
ð1Þ
where χeff(2) is the effective nonlinear susceptibility, and I(ωVIS) and I(ωIR) are the intensities of visible and IR input laser beams, respectively. The effective nonlinear susceptibility can be expressed as11 N χef f ð2Þ ¼
∑
R, β, γ
ÆMRβ Ay æ
εo ðωIR ωq iΓÞ
ð2Þ
where N is the density in unit volume, MRβ and Ay are the Raman and IR tensors, respectively, ÆMRβAyæ is the macroscopic average of molecular hyperpolarizability, εo is the dielectric constant of vacuum, ωq is the frequency of a normal vibration mode, and Γ is the damping constant. Equation 2 states that to generate SFG resonance, a vibration mode must be both Raman- and IR-active, and those vibration modes must be arranged without inversion symmetry in macroscopic space. Otherwise, χeff(2) is zero. Thus, the SFG signal is proportional to the square of the density of vibration modes arranged noncenstrosymmetrically within the laser probe volume. In cellulose Iβ, which is most abundant in higher plants, two adjacent glucosyl units in the cellulose chain are placed within a monoclinic unit cell with two-fold screw symmetry (P21 space group) in which certain functional groups of the glucose unit are noncentrosymmetrically ordered.12 The extension of this ordering along the crystalline microfibers allows for nonlinear optical activities.1315 In the cellulose Iβ structure, which is the dominant allomorph in secondary cell walls of higher plants, it is found that the CH2 group in the exocyclic 6CH2OH side chain and the intrachain-hydrogen-bonded OH groups are SFG-active in the CH and OH stretch vibration regions. Other CH and OH stretch vibrations in the crystalline cellulose are not detected in SFG because of the noncentrosymmetry rule of the SFG optical process. Likewise, all other amorphous carbohydrates and water molecules in the sample do not produce any SFG signal. This allows direct and selective detection of crystalline cellulose in lignocellulosic biomass without extraction of amorphous components from biomass or spectral peak deconvolution from amorphous components.
’ EXPERIMENTAL SECTION Materials. Commercially available Iβ-rich cellulose such as Avicel PH-101 (Fluka 11363), Sigmacell 20 (Sigma S3504), and IFC cotton linter (JustFiber C10CL FCC) were used in this study. Xylan (Sigma X4252) and alkali lignin (Aldrich 370967) were also used as hemicellulose and lignin, respectively. Wood chip samples were collected from tree trunk portions of Scarlet oak (Quercus coccinea), Scandinavian birch (Betula pendula), and Loblolly pine (Pinus taeda) and tested to demonstrate the selective detection of cellulose from woody cells. The
Figure 1. Schematic diagram of the SFG system used in this experiment. chemical compositions of wood samples were analyzed and are shown in Table S-1 of the Supporting Information. SFG Spectroscopy. A schematic view of the EKSPLA SFG system used in this experiment is given in Figure 1. The SFG spectrometer consisted of an optical parametric generator/amplifier (OPG/OPA) pumped by a Nd/Yag laser at a 10 Hz repetition rate. The laser pulse width was 27 ps. The OPG/OPA unit generated the tunable IR beam between 2 and 10 μm. The second harmonic output of the Nd:Yag laser was used for the visible light at 532 nm. The incident angles of the IR and visible beams were 56 (θIR) and 60 (θVIS) from the surface normal, respectively. The IR and visible beams were spatially and temporally overlapped at the surfaces of compressed powder pellets and wood chips, and the SFG signal was collected in a reflection geometry. A beam collimator was used to increase the collection efficiency of the SFG signal from the sample. The generated sum frequency signal was filtered using a monochromator and detected with a photomultiplier tube. All SFG spectra were obtained with all polarization for SFG output photons and s-polarized visible and p-polarized IR input laser beams. The SFG peak intensity was normalized with the visible and IR input layer intensities (eq 1). Fourier Transform Infrared and Raman Spectroscopy. The Fourier transform infrared (FT-IR) spectroscopy measurements were performed using a Thermo-Nicolet 760 with a DGTS detector. The FTIR spectrum of the cellulose and biomass samples was analyzed from 4000 to 400 cm1 to encompass the OH and CH stretching vibration regions as well as the OH and CH bending vibration regions. The data points were taken at 2 cm1 steps, and each spectrum shown here was the average of 100 spectra. All data were collected in the transmission mode. For the IR measurement, the powder samples were combined with KBr in a 1:10 ratio by weight. The Fourier transform Raman (FT-Raman) spectroscopy measurements were performed using a Renishaw Invia reflex single grating deep UV MicroRaman spectrometer with a 514 nm excitation source. The FT-Raman spectrum was obtained from 800 to 3800 cm1 with a 1 cm1 step.
’ RESULTS AND DISCUSSION Figure 2 compares the SFG spectrum of Avicel PH-101, which is one of the most widely used model cellulose samples in the literature, pressed into a pellet with its IR and Raman spectra. In IR and Raman, the νOH region (32003750 cm1) is extremely broad because of extensive hydrogen bonds within and between cellulose chains as well as the presence of water remaining in the amorphous region of the sample. The latter is problematic because it makes it extremely difficult to get meaningful structural information from the OH stretching vibrations of the cellulose. The CH stretching vibration region of the IR 2435
dx.doi.org/10.1021/bm200518n |Biomacromolecules 2011, 12, 2434–2439
Biomacromolecules
Figure 2. Comparison of IR, SFG, and Raman spectra of microcrystalline cellulose (Avicel PH-101) taken in ambient air.
Figure 3. Schematic view of two glucopyranose units in cellulose chain. The SFG-active CH2 and OH stretch peaks are highlighted with green and pink, respectively.
and Raman spectra has a broad peak centered at 2902 cm 1 in IR and 2898 cm1 in Raman and small shoulders at ∼2945 and ∼2965 cm1. In IR, an additional shoulder can be seen at ∼2860 cm1 . The strongest peak at ∼2900 cm1 is attributed to the stretch vibration of the CH groups at 1, 2, 3, 4, and 5 positions the glucopyranose unit (Figure 3). Its intensity is the largest in the CH stretch region because they are most abundant in the cellulose unit cell structure.16 The CH stretch peak of crystalline cellulose is somewhat broader than that of typical organic molecules in gas or liquid states probably because of some strains in the six-membered ring structure in the crystalline cellulose structure.17 The 4001700 cm1 region show richer vibration peaks, but its interpretation or assignment to specific vibration modes is difficult because multiple CC and CO stretching modes or CCH, OCH, and COH bending modes are involved in these vibration peaks.16
COMMUNICATION
The SFG spectrum of cellulose shows much sharper and fewer vibration peaks compared with IR and Raman, owing to the noncentrosymmetry selection rule of the SFG optical process. The absence of the peak at 2900 cm1 in SFG is due to the symmetry cancellation of the axial CH bonds. In the β-1,4linked glucopyranose unit present in the cellulose Iβ crystal shown in Figure 3, the axial CH bonds at the one, three, and five positions in the left glucopyranose unit are pointing upward, whereas those in the right unit are pointing downward. In contrast, the axial CH bonds at the two and four positions are downward in the left unit, whereas those are upward in the right unit. In this configuration, their CH stretch signal (2900 cm1) will be canceled because of inversion symmetry in SFG, although they are active in both IR and Raman (eq 2). This is similar to the inversion symmetry cancellation of the CH2 peaks in highly ordered self-assembled monolayers.1820 The peaks at 2850 (weak) and 2945 cm1 (strong) are attributed to the CH2 symmetric and asymmetric vibrations, respectively, of the exocyclic 6CH2OH group. These peak positions are consistent with those of ethyleneglycol.2123 The 3325 cm1 peak position in SFG is very close to the stretch peak of 3OH hydrogen-bonded to 5O and 2OH hydrogen-bonded to 6O in IR.24 Although these OH groups are positioned at the opposite sides of the glycosidic linkage, they are not symmetrically arranged because they are interacting with different oxygen atoms (2OH 3 3 6O and 3OH 3 3 5O). All 2 OH 3 3 6O and 3OH 3 3 5O groups are aligned in the same direction along the chain because the cellulose chains are aligned parallel in the cellulose Iβ crystal. In addition, the 2OH 3 3 6O group is more disordered than the 3OH 3 3 5O group.25 Therefore, these two peaks will not cancel each other in the cellulose Iβ crystal and will be detected in SFG. In IR, the OH stretching vibration peak with the transition dipole moment parallel to the cellulose chain direction appears at 3340 cm1, which is very close to the OH peak observed in the SFG spectra at 3325 cm1.24 SFG signals can also be generated at interfaces because the surfaces can naturally provide the noninversion symmetry. In fact, SFG is a very well known technique to be useful for buried interface studies.10 Several control experiments were carried out to confirm that the SFG signals of cellulose originate from the bulk not from the crystal surface. The OH SFG signal in the 32003400 cm1 region does not depend on the modification of the hydroxyl group at the cellulose crystallite surface. It is well known that the OH groups at the cellulose crystal surface as well as those in the amorphous phase can be deuterated via simply immersing in D2O through hydrogen/deuterium (H/D) exchange reactions.26 After proper H/D exchange treatment, there is neither discernible OD peak growth nor OH peak decrease in SFG, as shown in Figure 4a, although IR shows a large peak in the OD vibration region and a significant reduction of the OH vibration peak.27 It is also known that the cellulose whiskers prepared with H2SO4 and HCl have different surface functional groups.28 The H2SO4-treated whiskers contain negatively charged sulfate groups at the surface OH sites, whereas the HCl-treated whiskers keep the neutral OH group at the surface, but these two samples show the same SFG spectra in Figure 4b. These results indicate that the SFG signal of cellulose originates from the bulk structure of the cellulose crystallites, not from the surface of individual crystallites. The pressed pellet surface is not optically flat and the SFG signal is detected even if the sample surface is tilted (5 away 2436
dx.doi.org/10.1021/bm200518n |Biomacromolecules 2011, 12, 2434–2439
Biomacromolecules
COMMUNICATION
Figure 4. (a) SFG spectra of cellulose before and after D2O exchange. (b) SFG spectra of cellulose whiskers prepared from cotton using H2SO4 and HCl.
Figure 5. Comparison of (a) IR and (b) SFG spectra of xyloglucan, xylan, methylcarboxy cellulose, and lignin taken in humid ambient air.
from the laser reflection angle (Supporting Information). If SFG signal is generated only from the external surface of the pressed pellet, then the signal should be detected only at a small window of reflection angle satisfying the phase match condition, ωSFG 3 sin(θSFG) = ωVIS 3 sin(θVIS) þ ωIR 3 sin(θIR), where θ = beam propagation angle with respect to the normal of the reflection plane (Figure 1).10 It is believed that the detection of SFG signal from the sample consisting of randomly oriented crystalline domains is similar to the powder XRD principle. Although the incidence angles for IR and visible photons are fixed in the laboratory coordinate, their incidence angles with respect to individual crystalline cellulose domains vary depending on the orientation of the crystal within the probe volume, which would be roughly the irradiated area times the IR penetration depth. This is schematically illustrated in the Supporting Information. Thus, the SFG signal from randomly oriented cellulose samples will have a broad angular distribution. Another important and advantageous feature of SFG detection of the crystalline cellulose is the absence of any background contribution from the amorphous components such as hemicellulose and lignin in biomass samples. The chemical structures of the constituent monomeric units are similar in hemicellulose (such as xyloglucan and xylan) and cellulose; thus, the IR spectra of these samples resemble each other (Figure 5a). The spectral differences are so small that it is difficult to distinguish these
components in the mixture samples. In contrast, hemicellulose does not produce any SFG signals, as shown in Figure 5b, because they are amorphous. The side chains in xyloglucan prevent the crystalline ordering of the backbone, although it has the same β-1,4-D-glucopyranose backbone. Similarly, when the cellulose crystalline structure is disrupted, for example, by derivatization with carboxymethyl group, the SFG signal disappears as shown in Figure 5b. Lignin, another main component in lignocellulosic biomass, is cross-linked aromatic hydrocarbons. Again, it does not produce SFG signal at all because it is amorphous (Figure 5b). The fact that hemicellulose and lignin are SFG-inactive and only crystalline cellulose is SFG-active allows direct and selective detection of crystalline cellulose inside plant cell walls without any extraction or separation of amorphous components. Figure 6 compares the CH and OH stretching region of the IR and SFG spectra of woody cells of oak, birch, and pine trees and a physical mixture of 40% cellulose (Avicel), 30% xylan, and 30% lignin powders. The mixture composition was chosen to mimic typical hardwood compositions. The data presented in Figure 6 clearly demonstrate that SFG selectively detects the crystalline Iβ cellulose in biomass without any interference from hemicellulose or lignin components. The SFG signal intensities measured from wood samples are comparable to that of the physical mixture containing 40% Avicel. Because the SFG signal intensity is 2437
dx.doi.org/10.1021/bm200518n |Biomacromolecules 2011, 12, 2434–2439
Biomacromolecules
COMMUNICATION
Figure 6. (a) IR and (b) SFG spectra of pine, oak, birch, and a mixture of 40% Avicel, 30% xylan, and 30% lignin taken in humid ambient air.
governed by the number density of the cohesively ordered noncentrosymmetric vibration groups within the laser probe volume (eqs 1 and 2), the comparable SFG intensities for these samples imply that the crystalline nature of the cellulose in these wood samples could be similar to that of the Avicel. Further studies are under way to elucidate crystalline cellulose structure in biomass using SFG spectroscopy. These include quantitative relationships between SFG intensity and cellulose crystallinity in biomass, distinction of cellulose polymorphs (IR versus Iβ, II, and III), cellulose packing and orientation in primary and secondary cell walls, and interactions between cellulose and other cell matrix polymers.
’ CONCLUSIONS It is demonstrated that the crystalline cellulose inside lignocellulosic biomass can be selectively detected with SFG vibration spectroscopy. The SFG signals originate from the noncentrosymmetric ordering of the glucosyl group in the cellulose crystal. The selective and sensitive detection of crystalline cellulose in biomass with SFG can provide structural information that cannot be obtained with other conventional techniques such as XRD, solid-state NMR, IR, and Raman analyses because of interference from noncellulose constituents, large amorphous backgrounds, or overlapping peaks. Therefore, the SFG analysis presents a unique opportunity to reveal the crystalline properties of natural cellulose in lignocellulosic biomass without any separation process of noncellulose matrix. ’ ASSOCIATED CONTENT
bS
Supporting Information. Experimental details, wood composition analysis, SFG spectra of cotton linter and Sigmacell, nanowhiskers, and tilted samples, and polarization dependence of SFG intensity. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT This work was made possible through the support from the Penn State Center for Optical Technologies for SFG instrumentation and from North Carolina State University for Faculty Research & Professional Development Fund. This work was also
supported as part of The Center for LignoCellulose Structure and Formation, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under award number DE-SC0001090. We also acknowledge S. Denev and V. Gopalan for help with SFG system operations and maintenance, D. Cosgrove, T. Fitzgibbons, and J. Badding for Raman measurements, and J. Catchmark and T. Richard for some cellulose samples.
’ REFERENCES (1) Atalla, R. H.; Vanderhart, D. L. Native Cellulose - A Composite of Two Distinct Crystalline Forms. Science 1984, 223, 283–285. (2) Habibi, Y.; Lucia, L. A.; Rojas, O. J. Cellulose Nanocrystals: Chemistry, Self-Assembly, and Applications. Chem. Rev. 2010, 110, 3479–3500. (3) Park, S.; Baker, J. O.; Himmel, M. E.; Parilla, P. A.; Johnson, D. K. Cellulose Crystallinity Index: Measurement Techniques and Their Impact on Interpreting Cellulase Performance. Biotechnol. Biofuels 2010, 3, 10. (4) Park, S.; Johnson, D. K.; Ishizawa, C. I.; Parilla, P. A.; Davis, M. F. Measuring the Crystallinity Index of Cellulose by Solid State C-13 Nuclear Magnetic Resonance. Cellulose 2009, 16, 641–647. (5) Thygesen, A.; Oddershede, J.; Lilholt, H.; Thomsen, A. B.; Stahl, K. on the Determination of Crystallinity and Cellulose Content in Plant Fibres. Cellulose 2005, 12, 563–576. (6) Driemeier, C.; Calligaris, G. A. Theoretical and Experimental Developments for Accurate Determination of Crystallinity of Cellulose I Materials. J. Appl. Crystallogr. 2011, 44, 184–192. (7) Liitia, T.; Maunu, S. L.; Hortling, B.; Tamminen, T.; Pekkala, O.; Varhimo, A. Cellulose Crystallinity and Ordering of Hemicelluloses in Pine and Birch Pulps As Revealed by Solid-State NMR Spectroscopic Methods. Cellulose 2003, 10, 307–316. (8) Newman, R. H. Homogeneity in Cellulose Crystallinity between Samples of Pinus radiata Wood. Holzforschung 2004, 58, 91–96. (9) De Souza, I. J.; Bouchard, J.; Methot, M.; Berry, R.; Argyropoulos, D. S. Carbohydrates in Oxygen Delignification. Part I: Changes in Cellulose Crystallinity. J. Pulp Paper Sci. 2002, 28, 167–170. (10) Shen, Y. R. Surface-Properties Probed by Second-Harmonic and Sum-Frequency Generation. Nature 1989, 337, 519–525. (11) Vidal, F.; Tadjeddline, A. Sum-Frequency Generation Spectroscopy of Interfaces. Rep. Prog. Phys. 2005, 68, 1095–1127. (12) Nishiyama, Y.; Langan, P.; Chanzy, H. Crystal Structure and Hydrogen-Bonding System in Cellulose Iβ from Synchrotron X-ray and Neutron Fiber Diffraction. J. Am. Chem. Soc. 2002, 124, 9074–9082. (13) Brown, R. M.; Millard, A. C.; Campagnola, P. J. Macromolecular Structure of Cellulose Studied by Second-Harmonic Generation Imaging Microscopy. Opt. Lett. 2003, 28, 2207–2209. (14) Chu, S. W.; Chen, I. H.; Liu, T. M.; Sun, C. K.; Lee, S. P.; Lin, B. L.; Cheng, P. C.; Kuo, M. X.; Lin, D. J.; Liu, H. L. Nonlinear 2438
dx.doi.org/10.1021/bm200518n |Biomacromolecules 2011, 12, 2434–2439
Biomacromolecules
COMMUNICATION
Bio-Photonic Crystal Effects Revealed with Multimodal Nonlinear Microscopy. J. Microsc. 2002, 208, 190–200. (15) Mizutani, G.; Koyama, T.; Tomizawa, S.; Sano, H. Distinction between Some Saccharides in Scattered Optical Sum Frequency Intensity Images. Spectrochim. Acta, Part A 2005, 62, 845–849. (16) Cael, J. J.; Gardner, K. H.; Koenig, J. L.; Blackwell, J. Infrared and Raman-Spectroscopy of Carbohydrates. 0.5. Normal Coordinate Analysis of Cellulose 1. J. Chem. Phys. 1975, 62, 1145–1153. (17) Nishiyama, Y.; Johnson, G. P.; French, A. D.; Forsyth, V. T.; Langan, P. Neutron Crystallography, Molecular Dynamics, and Quantum Mechanics Studies of the Nature of Hydrogen Bonding in Cellulose Iβ. Biomacromolecules 2008, 9, 3133–3140. (18) Belkin, M. A.; Kulakov, T. A.; Ernst, K. H.; Yan, L.; Shen, Y. R. Sum-Frequency Vibrational Spectroscopy on Chiral Liquids: A Novel Technique to Probe Molecular Chirality. Phys. Rev. Lett. 2000, 85, 4474–4477. (19) Held, H.; Lvovsky, A. I.; Wei, X.; Shen, Y. R. Bulk Contribution from Isotropic Media in Surface Sum-Frequency Generation. Phys. Rev. B 2002, 66, 205110. (20) Weeraman, C.; Yatawara, A. K.; Bordenyuk, A. N.; Benderskii, A. V. Effect of Nanoscale Geometry on Molecular Conformation: Vibrational Sum-Frequency Generation of Alkanethiols on Gold Nanoparticles. J. Am. Chem. Soc. 2006, 128, 14244–14245. (21) Dreesen, L.; Humbert, C.; Hollander, P.; Mani, A. A.; Ataka, K.; Thiry, P. A.; Peremans, A. Study of the Water/Poly(ethylene glycol) Interface by IR-Visible Sum-Frequency Generation Spectroscopy. Chem. Phys. Lett. 2001, 333, 327–331. (22) Hommel, E. L.; Merle, J. K.; Ma, G.; Hadad, C. M.; Allen, H. C. Spectroscopic and Computational Studies of Aqueous Ethylene Glycol Solution Surfaces. J. Phys. Chem. B 2005, 109, 811–818. (23) Matsuura, H.; Miyazawa, T. Infrared Spectra and Molecular Vibrations of Ethylene Glycol and Deuterated Derivatives. Bull. Chem. Soc. Jpn. 1967, 40, 85–94. (24) Marechal, Y.; Chanzy, H. The Hydrogen Bond Network in Iβ Cellulose As Observed by Infrared Spectrometry. J. Mol. Struct. 2000, 523, 183–196. (25) Kono, H.; Yunoki, S.; Shikano, T.; Fujiwara, M.; Erata, T.; Takai, M. CP/MAS 13C NMR Study of Cellulose and Cellulose Derivatives. 1. Complete Assignment of the CP/MAS 13C NMR Spectrum of the Native Cellulose. J. Am. Chem. Soc. 2002, 124, 7506–7511. (26) Jeffries, R. The Amorphous Fraction of Cellulose and Its Relation to Moisture Sorption. J. Appl. Polym. Sci. 1964, 8, 1213–1220. (27) Horikawa, Y.; Sugiyama, J. Accessibility and Size of Valonia Cellulose Microfibril Studied by Combined Deuteration/Rehydrogenation and FTIR Technique. Cellulose 2008, 15, 419–424. (28) Roman, M.; Winter, W. T. Effect of Sulfate Groups from Sulfuric Acid Hydrolysis on the Thermal Degradation Behavior of Bacterial Cellulose. Biomacromolecules 2004, 5, 1671–1677.
2439
dx.doi.org/10.1021/bm200518n |Biomacromolecules 2011, 12, 2434–2439
