Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Chapter 3 Towards an Understanding of Cellulose Microfibril Dimensions from TEMPO-Oxidized Pulp Fiber Zehan Li,1 Noppadon Sathitsuksanoh,2 Wei Zhang,1 Barry Goodell,3 and Scott Renneckar*,4 1Department of Sustainable Biomaterials, Virginia Tech, 230 Cheatham Hall, Blacksburg Virginia 24061, United States 2Department of Chemical Engineering, Ernst Hall, Room 216, University of Louisville, Louisville, Kentucky 40292, United States 3Department of Microbiology, University of Massachusetts, 639 North Pleasant Street, Amherst, Amherst, Maine 01003, United States 4Department of Wood Science, University of British Columbia, 2424 Main Mall, Vancouver, BC, Canada V6T 1Z4 *E-mail: [email protected] A unique molecularly thin nanocellulose (MT nanocellulose) structure –obtained by 2,2,6,6-tetramethylpiperidin- 1 oxyl (TEMPO)-oxidation and sonication– was examined by TEM and solid-state 13C NMR to advance the current understanding on the supramolecular structure of cellulose I microfibrils. The width distribution of the microfibril was determined from TEM images, and a holistic view of the microfibril cross section was developed by integrating the height distribution result from previous work using atomic force microscopy imaging. Systematic changes of NMR spectra upon oxidation and sonication treatments were observed and attributed to the corresponding changes to crystallinity, glycosidic linkage torsion angles, as well as C6 primary hydroxyl conformations. Lastly, current microfibril cross-section models were collectively reviewed and the 24-chain diamond model was identified as the most credible representation for the experimental data and the known constraints. © 2017 American Chemical Society Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Introduction Nanocellulose has emerged as a field of research interest, not only because of the extensive availability and sustainability of its precursor cellulose (1), but also due to the broad chemical modification range, excellent physical and mechanical properties, as well as the enormous potential applications related to nanotechnology (2, 3). However, the accurate knowledge of the geometry of nanocellulose is still poorly resolved in the literature. Lacking exact knowledge of the geometry can misrepresent modeling efforts when nanocellulose is used in composite materials. Moreover, cellulose source and isolation methods impact nanocellulose dimensions creating ambiguity around cellulose microfibril geometry in the native state along with the isolated counterpart. In 1954, Frey-Wyssling suggested a near rectangular cross-sectional arrangement of cellulose molecules to describe the organization of plant derived cellulose microfibrils, based on microscopy and XRD evidence (4). Today, wood-based cellulose microfibrils have been proposed having a cross section shape approximate to either a hexagon (5), a rectangle (5, 6), or an ellipse (7), based on direct microscopy observation or indirect evidence from its biosynthesis from hexameric terminal complex of particle “rosettes” (8, 9). Since the rosettes impart a six-fold symmetry (10, 11), to the cellulose microfibril, hexagonal and elliptical cross sectional arrangements would seem to best reflect synthesis via this process. Cellulose chain numbers contained in the microfibril are believed to be constant based on the assumption that one terminal complex of 6-subunit rosette extrudes one microfibril (5). Some authors have suggested a 36-chain packing scheme (12, 13), while others have raised doubt over whether this packing scheme is too large to fit experimental observations and they have suggested a 24-chain packing scheme instead (5). Current understanding of cellulose microfibril cross-sectional structure has been deduced from a combination of plant cellulose biosynthetic origins, crystal lattice dimensional measurements, microfibril dimensions, as well as assumptions about cross section shapes (5, 13–16). However, due to the uncertainties in both the data and assumptions, agreement on the arrangement of the microfibril cross section arrangement has yet to be reached. Another approach is to look at the deconstruction of the microfibrils, which has been done through acid hydrolysis of wood pulp, which degrades cellulose microfibrils into fragments. Another method to isolate microfibrils is through the select oxidation of cellulose structure using 2,2,6,6-tetramethylpiperidin- 1 oxyl (TEMPO). Isogai and coworkers developed a facile way of producing nanocellulose via TEMPO-oxidation and mechanical agitation to deconstruct pulp fibers (6). Several efforts focused on the isolation, characterization, of the fine structure of the oxidized cellulose microfibrils derived from native cellulose sources, as well as applications for these materials (6). Utilizing the method to understand the fundamental structure of cellulose, Okita et al. demonstrated that the degree of surface oxidation corresponded with the X-ray data for cellulose microfibril dimensions (17). This surface only microfibril oxidation stemmed from the limitations of the reactants to penetrate the surface of the microfibril, along with the lack of solubility of the partially oxidized cellulose chains to peel away 56 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 from the surface. Unfortunately, direct visualization of native fibril dimensions cannot be determined as mechanical agitation to isolate individual nanofibrils influences the dimensions of the resulting isolated nanocellulose. This result was exemplified through extended sonication of TEMPO-oxidization of softwood kraft pulp fibers. Li and Renneckar isolated cellulose mono- and bi-layer molecular sheets (i.e.: molecularly thin nanocellulose, or “MT nanocellulose”) where the thickness profile of the sheets were determined by atomic force microscopy (AFM) on atomically smooth surfaces (18). Further investigation with X-ray diffraction, Raman, and FTIR indicated that delamination occurred along the (200) plane in the cellulose Iβ crystalline structure (19). Following this method Su et al. used synchrotron X-ray diffraction and small angle X-ray scattering (SAXS) confirming thickness of mono and bi-layer sheets (20). An advantage of the SAXS study was that nanocellulose was measured in the aqueous state, limiting self-assembly, revealing sub nanometer dimensions of sonicated TEMPO cellulose. In contrast to Li and Renneckar, Su et al. showed X-ray data that highlighted changes in the (110) plane, which suggested breakage of the extensive hydrogen bonding network of the fibril (within sheet) instead of intersheet bond breakage. In both cases, molecular thin cellulose particles were isolated, however the fracture process would have been characteristically different. To date, height (thickness) and length profiles of MT nanocellulose have been examined using atomic force microscopy. The width profile was not examined using this method, because of the inherent tip convolution effect when using AFM. Hence, development of a holistic (3D) structure for cellulose molecular sheets has not been possible. Transmission electron microscopy (TEM) has been applied in prior research to characterize the nanocellulose width and length profiles because TEM permits accurate horizontal resolution (21–28). Several methods have been used to prepare nanocellulose materials for observation by TEM including hydrolysis (21), TEMPO-oxidation combined with sonication (22, 27), using different raw material sources (wood pulp (22, 25, 27), cotton, tunicin, Avicel (21)), with negative staining techniques typically used to enhance contrast (21, 27). A better understanding of the supramolecular structure of MT nanocellulose, and the multiple MT nanocellulose layers that are assembled into the cellulose microfibril, could potentially be obtained through the use of TEM. Another powerful method to study cellulose microfibrils is cross polarized magic angle spinning solid-state 13C nuclear magnetic resonance (CP-MAS SS NMR) spectroscopy (29–31). The NMR chemical shifts of cellulose material can be obtained readily by conventional solid state 13C NMR to evaluate conformation, hydrogen bonding, and molecular packing (32). There is a clear association between hydrogen bonding and chain packing, and conformational shifts for the C6 on glucopyranose ring (Figure 1A) in the cellulose structure (33). It has also been reported that the glycosidic bond conformation, which is defined by two glycosidic linkage torsion angles (34) Φ (O5′–C1′–O4–C4) and Ψ (C1′–O4–C4–C5) (Figure 1B), affects the cellulose supramolecular structure significantly and is related to C1 and C4 chemical shifts (35). This torsion angle impacts the spacing of the cellobiose repeat unit and would be heavily influenced by the constraints of the microfibril. 57 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Figure 1. Schematic drawings represent (A) three possible conformations at glucopyranose C6 position, (B) cellulose torsion angles Φ, Ψ, and Χ (O6′–C6′–C5′–C4′). In the present study, TEMPO-oxidized kraft pulp was ultrasonicated at different time intervals to produce MT nanocellulose. TEM and NMR were utilized to examine the MT nanocellulose to obtain additional information on its unique structure, including: width profile distribution, crystallinity, glycosidic linkage torsion angles, C6 primary hydroxyl group conformations, as well as how changes in these features occurred during sonication. The structural features of MT nanocellulose were then assessed with regard to prior knowledge of microfibril structure to describe the microfibril dimensions. Experimental Section Materials Never-dried kraft pulp from the southeastern USA of southern yellow pine softwood species (obtained from Weyerhaeuser Co., Ltd. with reported 88% brightness and DP ranging from 1600-1694), was used as the starting material. NaClO, NaBr, and 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) were obtained from Sigma Aldrich. Ultrapure water used all experiments was generated by Millipore Systems (Direct-Q 3UV), with a conductivity of 0.30 μs/cm and purity < 5 ppb (18, 19). 58 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. TEMPO-Mediated Oxidation The kraft pulp was oxidized following previously published techniques with NaClO at 5 mmol of NaClO per gram of dry fiber used to convert accessible primary alcohol groups to carboxyl groups (18). The final degree of oxidation, determined by conductometric titration (36), was 1.43 mmol/g of fiber, equivalent to a DS of 0.23 (23% of the total AGUs had their C6 primary hydroxyl group converted to a carboxyl group). Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Sonication Both kraft pulp and oxidized pulp were sonicated to generate nanocellulose fibrils. Sonication was conducted at five time intervals (5, 30, 60, 120, and 240 min), at 0.1% (w/w) concentration of fiber slurry, in a temperature controlled bath at 4°C. A 19 mm diameter medium intensity horn was used to sonicate the fibril suspension at 20 kHz (VC700, Sonics and Materials). The sonicated suspension was centrifuged at 4500 rcf for 15 min and the decanted transparent supernatant was stored for later processing and analysis by TEM and NMR (lyophilized). The complete sample set for TEM and NMR analysis included: kraft wood pulp (WP), 120 min sonicated kraft wood pulp (WP120), TEMPO-oxidized wood pulp (WT), TEMPO-oxidized wood pulp that undergone 5, 30, 60, 120, and 240 min sonication (WT 5, WT30, WT60, WT120, and WT240). TEM Analysis Cellulose suspensions of ca. 5x10-3% (w/w) concentration were first deposited onto Formvar TEM grids (400-mesh); incubating for 5 min before the suspensions were blotted with filter paper. The cellulose samples were then immersed for 2 minutes in a 2% uranyl acetate solution for negative staining before blotting with filter paper. The stained samples were then immediately observed under a ZEISS 10CA TEM, operating at 60 kV. TEM images were analyzed with NIS-Elements BR software for fibril width measurement and statistical analysis. Four hundred measurements were made from 10-15 images for each sonication time level, and the images were magnified with the assistance of a “zoom” tool in the software package to ensure accurate measurements to the nearest resolved pixel. In order to avoid errors induced from incidental twisting (abrupt narrow parts with color changes) of the fibrils, portions of the fibrils that appeared to narrow abruptly or that had abrupt color change were avoided, when taking measurements. (CP/MAS) Solid-State 13C NMR Experiment Cellulose lyophilized samples were ground (Wiley® Mini Mill mesh #60, 250 μm) and powdered samples were packed into rotors for NMR scanning. NMR spectra were obtained on a Bruker AVANCE DPX 300 instrument, operating at 75 MHz carbon frequency, with a 4 mm (o.d.) rotor sample spinning at 6.5 kHz, total contact time 1ms, 3 s relaxation delay, and at ambient temperature. The chemical 59 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. shift scale was calibrated relative to tetramethylsilane (TMS), with the CH highfield peak set at 29.5 ppm. NMR Crystallinity Index Evaluation Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Crystalline (cr) and non-crystalline (non-cr) contributions were determined by peak integration. For the non-crystalline region was delimited in the 86.5 - 80.6 ppm range, and the crystalline was delimited between 93 - 86.5 ppm (29). The Crystalline Index (CI) was then calculated using equation (1) Results and Discussion TEM Evaluation of Nanocellulose The nanocellulose in the TEM exists in long flat fibril form; the width of individual fibrils was relatively uniform with periodic regions that narrow abruptly, indicating possible twists in the structure (Figure 2A-C). The nanocellulose width distribution was obtained by plotting 400 width data points for each sonication time level (Figure 2D). For short sonication times, width distributions were between 2 and 14 nm. Compared to this short sonication time, the average width decreased and leveled off to approximately 4 nm after 60 min sonication, which suggests that 60 min sonication would be sufficient in isolating majority of the individual microfibrils. This result also reconfirms and extends the depth of Johnson’s and Saito’s investigations related to sonication (27, 37). Tests of statistical significance indicate that the width difference between the 5 min and 30 min sonication levels, and 30 min and 60 min levels are significant; whereas the difference between 60 min and 120 min levels, and 120 min and 240 min levels are statistically insignificant. The overall average width for extended sonication groups “60MinPlus” (60, 120, and 240 min combined) was 3.93 nm. Even though shorter sonication time resulted in longer distribution tails at upper end, the minimum widths all have a cutoff value around 2 nm regardless of sonication time, indicating that an approximate 3-chain sheet in the 200 plane of the unit cell (one chain is 0.82 nm across based on unit cell values) could be the smallest MT nanocellulose structure, which is consistent with the microfibril cross section model suggested in our previous study (19). As sonication time increased, the majority of the measurements, become clustered with the upper box plot value near 5nm at the higher end of the measurements. This result suggested the maximum width of individual sheet is around 5 nm, corresponding to approximate 6 cellulose chains connecting each other side-by-side (4.92 nm). The larger width values (5 nm and above) could be the manifestation of the “non-individualized” fiber bundles and/or sheets connected side-by-side via inter-chain hydrogen bonds (13, 38). However, these values may provide insight to the way the microfibrils aggregate within the cell wall. For extended sonication levels (60, 120, 240 min, collectively noted as “60MinPlus” group in Figure 2), the majority (75% percentile) of the width measurements are below 5 nm. This 60 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 data clearly illustrates that sonication energy can induce fragmentation of the larger aggregates into smaller sections, which correspond in lateral dimension to chain lengths of approximately 6 cellulose chains across in the 200 plane. Figure 2. TEM images of sonicated nanocellulose (A: 5 min; B: 60 min; C: 120 min) scale bar 100 nm, (representative twisting features were indicated with arrows) and (D) nanocellulose width distribution from TEM Note 25%, 50%, and 75% percentiles are denoted as the long horizontal bars, 1% and 99% percentiles are denoted as “*”, the mean is denoted as “□”, and the full range of the distribution is in between the short horizontal bars at both ends of the box-whisker plot. Integrating the TEM measurements with the previous AFM results on microfibril thickness (18), the isolated cellulose molecular sheets can be visualized as a long flat ribbon with a thickness of approximately 1 nm (18) or less with width ranges from 2-5 nm (TEM data), and a length scale ranging from hundreds of nanometers to several micrometers (18, 22). The periodic twist structure of cellulose microfibrils reviewed previously (5, 21, 39), was observed as a twisting feature within the TEM images. The structure appears as width variations along 61 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. individual microfibrils (39), and is similar to that which has been observed in other cellulose fibril structures (21). Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 NMR Results: Crystallinity, Molecular Conformation, Chain Conformation Figure 3 contains the (CP/MAS) 13C NMR spectra across all treatment levels, showing the variations between the kraft pulp, oxidized pulp, and oxidized and sonicated pulp. A carboxylate peak emerges in the spectrum ca. 175 ppm for the oxidized samples (Figure 3) (6). For 13C NMR spectra of cellulose, the relative intensities of the peaks correspond to the proportion of the specific carbons giving rise to them (31), hence the carboxylate peak intensity agree with the degree of oxidation. Since oxidation only occurs on the fibril surface and ~23% of the C6 were converted to carboxylate group under our experimental conditions (see experimental section for details), the peak intensity ratio between carboxylate and C6 should be near 23%. From the spectra, ratios of (ICarboxylate : IC6) are: 19%, 22%, 19%, and 23% for WT, WT30, WT60, and WT120, respectively; which is in agreement with the degree of oxidation determined previously by conductometric titration. Figure 3. (CP/MAS) 13C NMR spectra of WP, WT, WT30, WT60, and WT120. (A) expansion of the carboxylate group chemical shift region, indicating that i) carboxylate group peak emerges after oxidation, and ii) the carboxylate peak shifts ca. 1 ppm towards upfield upon sonication. (B) 55-125 ppm region. 62 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. With increased sonication levels the carboxylate peak has an upfield displacement indicating a slightly different environment for these carboxylate peaks. This change may relate to the increased gg conformation at C6 position as shown by the ~1 ppm upfield shift in the carboxylate peak region upon the sonication treatment. Since after sonication, surface chain proportion should increase from 40-50% to 80%; accordingly more C6 is exposed to the surface and converted to (possibly more preferred) gg conformation. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 NMR Crystallinity Determination Different methods have been developed to evaluate the crystallinity index of cellulose with (CP/MAS) 13C NMR: curve fitting (40), peak area separation (41), and chemometric analysis (i.e. principle component analysis) (42). All of these methods seem to have their own limitations: curve fitting methods are not very precise in terms of chemical shifts assignments associated with the disordered regions, plus the results are operator dependent and difficult to reproduce even when conducted by the same operator (42). Peak area separation tends to overestimate the narrower crystalline peaks (29); and chemometrical analysis is model-independent providing more consistent results (42), but requires large scan numbers and extended data collecting time. The peak area separation method was applied in this study to compare the differences introduced by different treatment levels. NMR crystallinity index (CI) calculation results for all levels are presented in Table 1, and Figure 4 demonstrates the calculation for WP level. The general trend for CI variation across all levels is similar to the results derived from XRD previously, and the major cause for a significant CI drop upon sonication is that the delamination effect has destroyed a portion of the crystalline structure (19). While this result is in contrast to Heux and Vignon when homogenizing sugar beet pulp (43), this data was supported by crystal size measurement using the Scherrer equation, where the (200) plane thickness was reduced by ~30% (19). With regard to the difference between the CI absolute value determined by “XRD height” and “NMR peak integration” (noted that NMR CI values are overall much smaller than XRD CI), Park et al. (41) have provided a comprehensive account, essentially stating that the XRD height method is a “time-saving empirical measure of relative crystallinity” and is likely to overestimate the crystalline portion. One striking result though is the increase in crystallinity after oxidation. A plausible explanation is that part of the amorphous hemicellulose in the kraft pulp (up to 25% (44)) has been removed during the oxidation and the follow-up purifying process, leading to an increase in the relative proportion of the crystalline cellulose, and hence the CI. Similar CI increases for cellulose I also was observed by XRD analysis (45). 63 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Table 1. Crystallinity index (CI) value calculated by peak integration method Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 CI WP WT WT30 WT60 WT120 39.67% 43.50% 31.33% 30.33% 30.00% Figure 4. (CP/MAS) 13C NMR spectrum of kraft wood pulp. Inset demonstrates the area based crystallinity index (CI) calculation: CI=C/(C+A) (41). NMR Implication on Molecular and Chain Conformation Changes under Sonication C1 and C4 Chemical Shift and Torsion Angles Φ and Ψ Figure 5 shows the peak shifts for C1 and C4 after oxidation and different sonication treatment levels. It should be noted that hemicelluloses impact the chemical shifts at C1 and C4 and this is noted in the figure for the WP original pulp sample (46). It is also clear that after oxidation at the conditions used there was a significant reduction of the hemicellulose contribution to the spectra (Figure 5). Kuramae et al. report that most of the neutral sugars are removed during oxidation, especially the mannan component with only a minimum residual xylan (47). Hence, caution should be noted for differences in chemical shifts between wood pulp and oxidized pulp arising from differences in composition. However, oxidized pulp and sonicated oxidized pulp should have similar compositions making the following noted differences arise from changes of the local environment of the cellulose chains in the microfibril. C1 peak positions do not show any displacement after oxidation, but exhibit 0.5 ppm 64 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 displacement towards upfield upon sonication. Similarly, C4 downfield peaks do not show any displacement after oxidation, but exhibit a shift of ~0.2 ppm towards upfield upon sonication. C4 upfield broad peaks are sharpened, and their relative intensities to the downfield sharp peaks increase upon sonication. The following four implications can be derived from the changes seen in the spectra: i) Surface chains have an elevated impact on NMR spectra upon sonication due to an increase in the proportional surface area. Both oxidation and sonication treatments affect the signal from the surface microfibril structure (19, 25, 48) according to the relationship between crystallite width and surface area chain proportion determined by Vietor et al. (35) Cellulose microfibrils from kraft pulp should have approximately 40~50% of their chains on the surface. Sonication will have a major delamination effect on these microfibrils with 60 min of sonication on oxidized fiber reducing the average fibril thickness to around 1 nm (19), which results in an overall proportion of surface chain increasing to ~80%, if there is no aggregation during freeze drying. Because the NMR spectra reflect the mean conformation of surface and core chains (35), the spectra of the sonicated samples are likely to be more representative of surface chains as opposed to the core chains. ii) Oxidation does not change torsion angles of glycosidic linkages but sonication does. C1 and C4 chemical shifts are dependent on Φ and Ψ (49, 50), although not in a linear fashion as occurs with Χ (33). Both WP and WT have C1 chemical shifts at 105 ppm (Figure 5) while the three sonication groups WT30, 60, and 120 all have their C1 chemical shifts displaced 0.5 ppm towards upfield to 104.5 ppm. This indicates that oxidation does not significantly affect the torsion angle Φ but sonication does. Unlike C1, the C4 chemical shifts do not reveal very explicit displacements across the five levels. The three sonication levels do have their downfield peaks blunted, indicating a tendency for upfield displacement, or change in Ψ. Therefore, both spectral changes at the C1 and C4 position indicate that oxidation does not change the glycosidic linkage conformation but sonication does cause a change in bond angles. This means the addition of carboxylate group on the surface has little impact on glycosidic linkage torsion angles, as long as the microfibril is intact; and once the microfibril is delaminated, glycosidic linkage will gain more freedom and exhibit more arrangements. This change in arrangement is important when evaluating the delamination mechanism of the microfibril, as changes to intersheet intermolecular bonds vs. intrasheet molecular hydrogen bonds possibly restrain the glycosidic bond in different fashion. iii) Glycosidic linkage changes are more pronounced on C1 than C4 chemical shifts. Since sonication alters torsion angles at the glycosidic bond, assuming the rotation of each linkage is shared almost equally between Φ and Ψ (35), the unbalanced response from C1 and C4 chemical shift displacements may well support Vietor’s theory that glycosidic conformation affects C1 chemical shifts more than C4 shifts, either through association with torsion angle or through hydrogen bonding at O3 (35). iv) Oxidation increases crystallinity, while sonication reduces crystallinity but increases the proportional area of fibril surface cellulose chains. Three features are identified contrasting the C4 downfield sharp peaks with upfield broad peaks: 1) relative intensity of the upfield peaks versus the downfield peaks increased significantly after sonication, indicating decreased crystallinity of the 65 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 sonicated groups; 2) the C4 upfield broad peaks sharpened at ca. 84 ppm upon sonication according to Newman’s assignment (51), which suggests an increased proportion of surface chains, and corroborates previous published results (18, 19); and 3) the C4 upfield broad peaks exhibit a downfield displacement from 83 ppm to 83.75 ppm while the WT C1 peak exhibits a receding slope from 100-102 ppm after oxidation. This evidence indicated the removal of residual hemicellulose during oxidation (47) which resulted in reduced hemicellulose peaks at ca. 81.9, 81.2 ppm and 101-102 ppm, according to previously published peak assignments (5, 52). Figure 5. (CP/MAS) 13C NMR spectra of the peak shifts for C1 and C4 under oxidation and different sonication treatment levels C6 chemical shift and torsion angle Χ. The peak shifts for C6 following oxidation and increasing sonication treatment (Figure 6) show that the relative intensities of the C6 downfield peak were reduced, while the upfield peak tended to shift to a lower ppm (63→2.5 ppm). These observations lead to the following three implications: i) gg conformation proportion rises with an increased proportion of surface chains. The C6 chemical shift was confirmed to have a linear relationship with torsion angle Χ’s three energy minimal positions (tg Χ=300°, gt Χ=180°, gg Χ=60°) (33, 53). A reduction in downfield peak intensity (65 ppm, related to tg conformation) and the upfield peak shift towards 62 ppm (related to gg conformation) (50, 53, 54) both indicate more CH2(OH) side groups were converted to gg conformation from the dominant tg and gt conformation (55, 56) when additional surface chains were exposed during sonication. ii) Surface chains favor gg conformation. During sonication, cellulose microfibrils become delaminated and the total 66 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 surface area therefore is expected to increase dramatically (18, 19). Enriched gg conformation therefore suggests the regioselectivity of the C6 primary hydroxyl group on the fibril surface, which would require that either surface chains are more energy favorable towards gg conformation than the inner chains, their spatial arrangements have an increased probability for gg conformation, or both. While Newman et al. have confirmed gg conformation for cellulose I at the cellulose-water interface (54), our results suggest that in the solid state too, the gg conformation might also be more favorable, and may even dominate, at the fibril surface. iii) Carboxylate side groups may contribute in adopting gg conformation. Because TEMPO-oxidation converts ~23% of the C6 primary hydroxyl groups to carboxyl groups, the larger carboxyl groups may also contribute to gg formation more as a stereochemically-preferred conformation compared to either gt or tg conformations at the cellulose-air interface. Interestingly, this trend is not apparent when contrasting WP and WT spectra, as the residual hemicellulose in WP may contribute to the upfield broad peak, which may then be large enough to disturb the trend. Figure 6. (CP/MAS) 13C NMR spectra of the peak shifts for C6 under oxidation and different sonication treatment levels. Reflections on Cellulose Microfibril Supramolecular Structure Many research efforts revolving around the cellulose Iβ microfibril structure of wood have been devoted into the understanding of its internal chain structure and size (5, 10, 11, 40, 57–61). Nevertheless, three key areas of understanding must be resolved before a well-defined microfibril structure is clearly identified: lateral dimensions, cross section shape, and chain packing numbers within individual microfibrils (5, 13, 39, 62). 67 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 Microfibril lateral dimensions have been characterized in many plant species using many techniques (SEM, TEM, AFM, XRD, SAXS, NMR) and it is understood that the lateral dimensions are source dependent but in a range from approximately 2-7 nm (7, 18). The cross sectional shape of plant cellulose microfibrils has been suggested as either hexagonal (13), rectangular (63), or approximate elliptical (7, 64), but recent crystallography data of microfibrils from lignified wood cells indicates that the rectangular shape is more likely to be the case (5). The chain packing numbers are determined by the number of cellulose synthase units found within the rosette terminal complex (TC) in the cell membrane of wood cells (11). The microfibril cross-sectional area and the chain packing patterns are used together to derive the chain packing numbers (14). The number of chains extruded from the TC are assumed to be constant under normal conditions (although an exception has also been reported) (5), hence the microfibril diameter (lateral dimensions) should also be consistent at least within the same plant species. A 36-chain model has frequently been suggested after Herth’s initial proposal based on estimation from electron microscopy (11, 12, 14, 15, 65), and this was deduced from the diameter of the rosette TC and the fact that each rosette is composed of 6 subunits in non-lignified plants (66). However, a recent study has challenged this model, arguing that the actual cross-sectional area (assuming a circular shape) of a microfibril can only accommodate 22 chains sufficiently. Thus, a 24-chain model has instead been proposed (5). Despite the exposure to pulping treatment (but not drying), the TEMPO oxidized fragmented fibril isolated in our current work should accurately represent the fibril in the native state. We argue that the width of the microfibril measured ranging from, 2-12 nm, with an average value of 3.93nm for the extended sonication time, should therefore be relatively close to that found in woody plant species. Using molecular dynamics analyses, Oheme et al. suggest that microfibrils aggregate together at the 110 plane via hydrogen bonds (67). This would provide justification for the widths we measured in the TEM converging on a value of 3.93 nm at the lower end. In height data measured previously, we reported less than 1% of our measurements were over 3.11 nm, with average values ranging from 1.38 to 0.74 nm depending upon the sonication time (18). Because of the fibril fragmentation, this data implies an upper bound of 3.11 nm. These data agrees with the 24-chain packing scheme recently proposed by Fernandes et al (5), as their two models for a diamond shape fibril and square shaped fibril are 3.2 x 3.9 nm and 3.2 x 3.1 nm, respectively. Based on the distribution of widths, the data supports the diamond shape, as width measurements were observed at 2 nm, which would account for delaminated sheets of the fibrils near the top and bottom of the diamond geometry. 68 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Conclusions 1. 2. 3. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 4. TEM analysis of sonicated TEMPO oxidized cellulose converged at width values of 3.93 nm for sonication times of 60 - 240 mins. Solid state NMR was used to estimate the oxidation level of cellulose, nearing the conductometric titration value which represents 23% of the primary hydroxyl groups. Crystallinity measurements and glycosidic torsion angles reveal disruption of the microfibril surface with sonication, while there was a minimum impact as a function of sonication time: 30, 60, 120 min. Experimental data fits the 24 chain diamond shaped microfibril model. Acknowledgments This work was funded in part by the Institute of Critical Technology and Applied Science of Virginia Tech, via the ICTAS Doctoral Scholar’s program (support for Zehan Li, formerly Qingqing Li). Additional support is from the Canada Research Chairs program for S. Renneckar’s Chair in Advanced Renewable Materials. B. Goodell was supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, the Center for Agriculture, Food and the Environment, and the Microbiology department at University of Massachusetts Amherst, under project number MAS00511. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the Canada Research Chairs, USDA or NIFA.” References 1. 2. 3. 4. 5. 6. Klemm, D.; Heublein, B.; Fink, H. P.; Bohn, A. Cellulose: Fascinating biopolymer and sustainable raw material. Angew. Chem., Int. Ed. 2005, 44, 3358–3393. Klemm, D.; Kramer, F.; Moritz, S.; Lindström, T.; Ankerfors, M.; Gray, D.; Dorris, A. Nanocelluloses: A new family of nature-based materials. Angew. Chem., Int. Ed. 2011, 50, 5438–5466. Renneckar, S.; Zink-Sharp, A.; Esker, A. R.; Johnson, R. K.; Glasser, W. G. Novel Methods for Interfacial Modification of Cellulose-Reinforced Composites. In Cellulose Nanocomposites: Processing, Characterization, and Properties; Oksman, K., Sain, M., Eds; ACS Symposium Series 938, American Chemical Society: Washington, DC, 2006; Chapter 7, pp 78−96. Frey-Wyssling, A. The fine structure of cellulose microfibrils. Science 1954, 80–2. Fernandes, A. N.; Thomas, L. H.; Altaner, C. M.; Callow, P.; Forsyth, V. T.; Apperley, D. C.; Kennedy, C. J.; Jarvis, M. C. Nanostructure of cellulose microfibrils in spruce wood. Proc. Natl. Acad. Sci. U.S.A. 2011, E1195–E1203. Isogai, A.; Saito, T.; Fukuzumi, H. TEMPO-oxidized cellulose nanofibers. Nanoscale 2011, 3, 71–85. 69 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. 7. 8. 9. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Leppänen, K.; Andersson, S.; Torkkeli, M.; Knaapila, M.; Kotelnikova, N.; Serimaa, R. Structure of cellulose and microcrystalline cellulose from various wood species, cotton and flax studied by X-ray scattering. Cellulose 2009, 16, 999–1015. Mueller, S. C.; Brown, R. M., Jr. Evidence for an intramembrane component associated with a cellulose microfibril-synthesizing complex in higher plants. J. Cell. Biol. 1980, 84, 315–26. Cosgrove, D. J. Growth of the plant cell wall. Nat. Rev. Mol. Cell Biol. 2005, 6, 850–861. Doblin, M. S.; Kurek, I.; Jacob-Wilk, D.; Delmer, D. P. Cellulose biosynthesis in plants: From genes to rosettes. Plant Cell Physiol. 2002, 43, 1407–1420. Brown, R. M.; Saxena, I. M. Cellulose biosynthesis: A model for understanding the assembly of biopolymers. Plant Physiol. Biochem. 2000, 38, 57–67. Endler, A.; Persson, S. Cellulose synthases and synthesis in Arabidopsis. Mol. Plant. 2011, 4, 199–211. Ding, S.; Himmel, M. E. The maize primary cell wall microfibril: A new model derived from direct visualization. J. Agric. Food Chem 2006, 54, 597–606. Saxena, I. M.; Brown, R. M. Cellulose biosynthesis: Current views and evolving concepts. Ann. Bot. 2005, 96, 9–21. Somerville, C.; Bauer, S.; Brininstool, G.; Facette, M.; Hamann, T.; Milne, J.; Osborne, E.; Paredez, A.; Persson, S.; Raab, T.; Vorwerk, S.; Youngs, H. Toward a systems approach to understanding plant-cell walls. Science 2004, 306, 2206–2211. Bessueille, L. A survey of cellulose biosynthesis in higher plants. Plant Biotech. 2008, 25, 315. Okita, Y.; Saito, T.; Isogai, A. Entire surface oxidation of various cellulose microfibrils by TEMPO-mediated oxidation. Biomacromolecules 2010, 6, 1696–1700. Li, Q.; Renneckar, S. Molecularly thin nanoparticles from cellulose: Isolation of sub-microfibrillar structures. Cellulose 2009, 16, 1025–1032. Li, Q.; Renneckar, S. Supramolecular structure characterization of molecularly thin cellulose I nanoparticles. Biomacromolecules 2011, 12, 650–659. Su, Y.; Burger, C.; Ma, H.; Chu, B.; Hsiao, B. S. Exploring the nature of cellulose microfibrils. Biomacromolecules 2015, 16, 1201–1209. Elazzouzi-Hafraoui, S.; Nishiyama, Y.; Putaux, J. L.; Heux, L.; Dubreuil, F.; Rochas, C. The shape and size distribution of crystalline nanoparticles prepared by acid hydrolysis of native cellulose. Biomacromolecules 2008, 9, 57–65. Shinoda, R.; Saito, T.; Okita, Y.; Isogai, A. Relationship between length and degree of polymerization of TEMPO-oxidized cellulose nanofibrils. Biomacromolecules 2012, 13, 842–849. Eichhorn, S. J. Cellulose nanowhiskers: promising materials for advanced applications. Soft Matter 2011, 7, 303–315. 70 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 24. Eichhorn, S. J.; Dufresne, A.; Aranguren, M.; Marcovich, N. E.; Capadona, J. R.; Rowan, S. J.; Weder, C.; Thielemans, W.; Roman, M.; Renneckar, S.; Gindl, W.; Veigel, S.; Keckes, J.; Yano, H.; Abe, K.; Nogi, M.; Nakagaito, A. N.; Mangalam, A.; Simonsen, J.; Benight, A. S.; Bismarck, A.; Berglund, L. A.; Peijs, T. Review: Current international research into cellulose nanofibres and nanocomposites. J. Mater. Sci. 2010, 45, 1–33. 25. Saito, T.; Hirota, M.; Tamura, N.; Kimura, S.; Fukuzumi, H.; Heux, L.; Isogai, A. Individualization of nano-sized plant cellulose fibrils by direct surface carboxylation using TEMPO catalyst under neutral conditions. Biomacromolecules 2009, 10, 1992–1996. 26. Saito, N.; Usui, Y.; Aoki, K.; Narita, N.; Shimizu, M.; Hara, K.; Ogiwara, N.; Nakamura, K.; Ishigaki, N.; Kato, H.; Taruta, S.; Endo, M. Carbon nanotubes: Biomaterial applications. Chem. Soc. Rev. 2009, 38, 1897–1903. 27. Johnson, R. K.; Zink-Sharp, A.; Renneckar, S.; Glasser, W. G. A new bio-based nanocomposite: Fibrillated TEMPO-oxidized celluloses in hydroxypropylcellulose matrix. Cellulose 2009, 16, 227–238. 28. Samir, M.; Alloin, F.; Dufresne, A. Review of recent research into cellulosic whiskers, their properties and their application in nanocomposite field. Biomacromolecules 2005, 6, 612–626. 29. Zuckerstaetter, G. The elucidation of cellulose supramolecular structure by 13C CP-MAS NMR. Lenzinger Ber. 2009, 87, 38. 30. Ibbett, R. N.; Domvoglou, D.; Fasching, M. Characterisation of the supramolecular structure of chemically and physically modified regenerated cellulosic fibres by means of high-resolution Carbon-13 solid-state NMR. Polymer 2007, 48, 1287–1296. 31. Atalla, R. H.; VanderHart, D. L. The role of solid-state carbon-13 NMR spectroscopy in studies of the nature of native celluloses. Solid State Nucl. Magn. Reson. 1999, 15, 1–19. 32. Atalla, R. H.; Isogai, A. Recent developments in spectroscopic and chemical characterization of cellulose. Polysaccharides 2005, 123–157. 33. Suzuki, S.; Horii, F.; Kurosu, H. Theoretical investigations of 13C chemical shifts in glucose, cellobiose, and native cellulose by quantum chemistry calculations. J. Mol. Struct. 2009, 921, 219–226. 34. French, A.; Johnson, G. What crystals of small analogs are trying to tell us about cellulose structure. Cellulose 2004, 11, 5–22. 35. Vietor, R. J.; Newman, R. H.; Ha, M.-A.; Apperley, D. C.; Jarvis, M. C. Conformational features of crystal-surface cellulose from higher plants. Plant J. 2002, 30, 721–731. 36. Katz, S.; Beatson, R. P.; Scallan, A. M. The determination of strong and weak acidic groups in sulfite pulps. Sven. Papperstidn. 1984, 87, R48–R53. 37. Saito, T.; Okita, Y.; Nge, T. T.; Sugiyama, J.; Isogai, A. TEMPO-mediated oxidation of native cellulose: Microscopic analysis of fibrous fractions in the oxidized products. Carbohydr. Polym. 2006, 65, 435–440. 38. Emons, A. M. C. Methods for visualizing cell-wall texture. Acta Bot. Neerl. 1988, 37, 31–38. 39. Atalla, R. H.; Brady, J. W.; Matthews, J. F.; Ding, S.-Y.; Himmel, M. E. Structures of Plant Cell Wall Celluloses. In Biomass Recalcitrance: 71 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. 40. 41. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. Deconstructing the Plant Cell Wall for Bioenergy; Himmel, M. E., Ed. Blackwell Publishing: Oxford, 2008; Vol. xviii. Hult, E. L.; Larsson, P. T.; Iversen, T. A comparative CP/MAS C-13-NMR study of the supermolecular structure of polysaccharides in sulphite and kraft pulps. Holzforschung 2002, 56, 179–184. Park, S.; Baker, J.; Himmel, M.; Parilla, P.; Johnson, D. Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance. Biotechnol. Biofuels 2010, 3 (1), 10. Rondeau-Mouro, C.; Bizot, H.; Bertrand, D. Chemometric analyses of the 1H-13C cross-polarization build-up of celluloses NMR spectra: A novel approach for characterizing the cellulose crystallites. Carbohydr. Polym. 2011, 84, 539–549. Heux, L.; Dinand, E.; Vignon, M. R. Structural aspects in ultrathin cellulose microfibrils followed by 13C CP-MAS NMR. Carbohydr. Polym. 1999, 40, 115–124. Sjöström, E., Wood Chemistry: Fundamentals and Applications, 2nd ed.; Academic Press: San Diego, CA, 1993; p 293. Saito, T.; Isogai, A. TEMPO-mediated oxidation of native cellulose. The effect of oxidation conditions on chemical and crystal structures of the waterinsoluble fractions. Biomacromolecules 2004, 5, 1983–1989. Maunu, S.; Liitiä, T.; Kauliomäki, S.; Hortling, B.; Sundquist, J. 13C CPMAS NMR investigations of cellulose polymorphs in different pulps. Cellulose 2000, 7, 147–159. Kuramae, R.; Saito, T.; Isogai, A. TEMPO-oxidized cellulose nanofibrils prepared from various plant holocelluloses. React. Funct. Polym. 2014, 85, 126–133. Okita, Y.; Saito, T.; Isogai, A. Entire surface oxidation of various cellulose microfibrils by TEMPO-mediated oxidation. Biomacromolecules 2010, 11, 1696–1700. Jarvis, M. C. Relationship of chemical shift to glycosidic conformation in the solid-state 13C NMR spectra of (1→4)-linked glucose polymers and oligomers: Anomeric and related effects. Carbohydr. Res. 1994, 259, 311–18. Horii, F.; Hirai, A.; Kitamaru, R. Cross-Polarization-Magic Angle Spinning Carbon-13 NMR Approach to the Structural Analysis of Cellulose. In The Structures of Cellulose: Characterization of the Solid States; Atalla, R. H., Ed.; ACS Symposium Series 340; American Chemical Society: Washington, DC: 1987; Chapter 6, pp 119−134. Newman, R. H. Evidence for assignment of C-13 NMR signals to cellulose crystallite surfaces in wood, pulp and isolated celluloses. Holzforschung 1998, 52, 157–159. Hult, E. L.; Larsson, P. T.; Iversen, T. A comparative CP/MAS C-13-NMR study of cellulose structure in spruce wood and kraft pulp. Cellulose 2000, 7, 35–55. Horii, F.; Hirai, A.; Kitamaru, R. Solid-state 13C-NMR study of conformations of oligosaccharides and cellulose. Polym Bull. 1983, 10, 357–361. 72 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017. Downloaded by CITY UNIV OF HONG KONG on October 25, 2017 | http://pubs.acs.org Publication Date (Web): October 23, 2017 | doi: 10.1021/bk-2017-1251.ch003 54. Newman, R.; Davidson, T. Molecular conformations at the cellulose–water interface. Cellulose 2004, 11, 23–32. 55. Maréchal, Y.; Chanzy, H. The hydrogen bond network in I[beta] cellulose as observed by infrared spectrometry. J. Molec. Struct. 2000, 523, 183–196. 56. Kondo, T. Hydrogen Bonds in Cellulose and Cellulose Derivatives. In Polysaccharides: Structural Diversity and Functional Versatility, 2nd ed.; Dumitriu, S., Ed.; Marcel Dekker: New York, 2005; pp 69−98. 57. Fengel, D. Ultrastructural behavior of cell wall polysaccharides. Tappi 1970, 53, 7. 58. Brown, R. M. Cellulose structure and biosynthesis: What is in store for the 21st century? J. Polym. Sci. Part A: Polym. Chem. 2004, 42, 487–495. 59. Osullivan, A. C. Cellulose: The structure slowly unravels. Cellulose 1997, 4, 173–207. 60. Hult, E. L.; Iversen, T.; Sugiyama, J. Characterization of the supermolecular structure of cellulose in wood pulp fibres. Cellulose 2003, 10, 103–110. 61. Jakob, H. F.; Fengel, D.; Tschegg, S. E.; Fratzl, P. The elementary cellulose fibril in Picea abies: Comparison of transmission electron microscopy, small-angle X-ray scattering, and wide-angle X-ray scattering results. Macromolecules 1995, 28, 8782–8787. 62. Nishiyama, Y. Structure and properties of the cellulose microfibril. J. Wood Sci. 2009, 55, 241–249. 63. Frey-Wyssling, A.; Muehlenthaler, K. The elementary fibrils of cellulose. Makromol. Chem. 1963, 62, 25–30. 64. Beck-Candanedo, S.; Roman, M.; Gray, D. G. Effect of reaction conditions on the properties and behavior of wood cellulose nanocrystal suspensions. Biomacromolecules 2005, 6, 1048–1054. 65. Somerville, C. Cellulose synthesis in higher plants. Annu. Rev. Cell Dev. Biol. 2006, 22, 53–78. 66. Herth, W. Arrays of plasma-membrane “rosettes” involved in cellulose microfibril formation of Spirogyra. Planta 1983, 159, 347–356. 67. Oehme, D. P.; Doblin, M. S.; Wagner, J.; Bacic, A.; Downton, M. T.; Gidley, M. J. Gaining insight into cell wall cellulose macrofibril organisation by simulating microfibril adsorption. Cellulose 2015, 22, 3501–3520. 73 Agarwal et al.; Nanocelluloses: Their Preparation, Properties, and Applications ACS Symposium Series; American Chemical Society: Washington, DC, 2017.