One-pot tandem reactions for the preparation of esterified cellulose nanocrystals with 4-dimethylaminopyridine as a catalyst

Abstract

An efficient approach for the manufacture of esterified cellulose nanocrystals (E-CNCs) via one-pot tandem reactions with 4-dimethylaminopyridine (DMAP) as a catalyst under mild operating conditions is put forward. The effects of ball milling time, reaction temperature and ultrasonication time on the yield and degree of substitution (DS) are explored. Characterization indicates the successful esterification of the hydroxyl groups of cellulose. The micromorphology and microstructure of the prepared E-CNCs are studied using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Results show that the E-CNCs are short rod-like particles 130–230 nm in length and 20–40 nm in width, forming an interconnected network structure. X-ray diffraction (XRD) results indicate that the crystallinity index increases from 63.5% to 77.2%. The thermal properties of the E-CNCs are investigated by thermogravimetric analysis (TGA) and the results show that the E-CNCs exhibit higher thermal stability than cellulose pulp.

---

1. Introduction

Cellulose nanocrystals (CNCs) have been used as reinforcements in polymers due to their high tensile strength, high flexibility and good dynamic mechanical properties.¹ However, as for highly polar CNCs, the main challenge is related to their uniform dispersion in non-aqueous mediums or in a polymeric matrix, and the difficulty in incorporating them into the most common apolar polymers.² In order to improve the dispersion, the surface characteristics of CNCs must be changed from hydrophilic to more hydrophobic. Acetylation is an efficient chemical modification of the hydroxyl groups on cellulose, resulting in a more hydrophobic surface. Previously reported studies on surface-modified cellulose nanofibers have focused on modification of isolated nanofibers, which usually involves fussy and time-consuming multi-step synthetic protocols.³

Nowadays, one-pot reactions have become one of the most attractive synthetic methodologies for reasons of their energy efficiency and general environmental friendliness.⁴ The one-pot reaction process has been proven to have several advantages over step-wise operations, as it avoids the isolation of intermediate species, thereby considerably reducing waste generation, increasing efficiency and minimizing the use of solvents, reagents, time and energy.⁵ Moreover, it was also found that in most cases the overall yields of one-pot processes are usually higher than those obtained from the corresponding step-wise operations. Since the middle of the 1960s, 4-dimethylaminopyridine (DMAP) has been widely used in organic synthesis as a catalyst, for example in Michael additions, aldol reactions and esterification reactions.^6–8 Attracted to the efficiency of the organo-catalyst and the advantage of using water as solvent, we chose DMAP as the catalyst in this work.

Mechanochemistry combines mechanical phenomena and chemical phenomena on a molecular scale, involving phase transition, size reduction and polymer degradation with the effects of compression and friction – cavitation-related phenomena.⁹ Size reduction of the solid and accretion of its specific surface area during mechanical milling are accompanied by distortion of chemical bonds and extension of bond lengths, due to the imposed stress, and when the imposed stress is beyond the chemical bonding energy bond rupture occurs, which effectively activates the functional chemical groups.¹⁰ Therefore, the mechanochemistry technique would be a potential method to manufacture surface-modified cellulose fillers under mild conditions.

Herein, we gain deeper insight into the efficiency of the nanocrystallization and acetylation of cellulose, in particular, disclosing a convenient and versatile method to manufacture functionalized E-CNCs in conjunction with ball milling and ultrasonication in tandem, in which the nanocrystallization and acetylation of cellulose take place simultaneously. Furthermore, the morphology, structure, spectroscopic and thermal properties of the esterified cellulose nanocrystals are also investigated.

2. Experimental

2.1 Materials

Bamboo pulp was supplied by Nanping Paper Co., Ltd. (Nanping, Fujian, China), and acetic acid and 4-dimethylaminopyridine (DMAP) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). All chemicals used in this work were of analytical grade and were used without any further purification.

2.2 Manufacture of esterified cellulose nanocrystals

The bamboo pulp was cut into pieces and beaten to form cellulose pulp with a Fiber Standard Dissociation device (GBT-A, Changchun Yueming Small Testing Machine Co., Ltd., China) for 30 min at 2000 rpm. Mechanochemical pretreatment of the cellulose pulp was performed within a planetary ball mill equipped with four agate jars. A mixture of 1 g cellulose pulp, 42 g acetic acid, 0.02 g DMAP and twenty 5 mm agate balls was added into the agate jar and milled at a rotational speed of 500 rpm for 1, 1.5, 2, 2.5 and 3 h. The samples were labeled as BM (1 h), BM (1.5 h), BM (2 h), BM (2.5 h) and BM (3 h), respectively. After milling, the samples were introduced into a 200 mL two-necked round-bottomed flask equipped with a condenser and a polytetrafluoroethylene (PTFE)-coated stirring device, and kept at 100–140 °C in an oil bath for 5 h under continuous stirring, followed by ultrasonication treatment at 20 kHz for 2–4 h in an ultrasonic reactor JY98-IIIN (Ning Bo Scientz Biotechnology Co., Ltd., China). After the reaction, the mixture was washed to neutralise it with double distilled water by repetitive centrifugations at 10 000 rpm for 15 min, and then the esterified cellulose nanocrystals were obtained. A schematic representation of the manufacture of the E-CNCs is shown in Fig. 1, and the mechanism of formation of the E-CNCs is outlined in Scheme 1.

Fig. 1 Schematic model of the manufacture of E-CNCs via one-pot tandem reactions.

Scheme 1 Mechanism of the formation of E-CNCs with DMAP as a catalyst.

2.3 Determination of yield and degree of substitution (DS) of E-CNCs

The total volume of the manufactured E-CNC suspension was measured. A specified amount of the E-CNC suspension was then transferred to a weighing bottle, followed by freeze-drying. The resulting sample was weighed. The final yield was obtained from the average of three runs of measurements and calculated according to eqn (1):where m₁ is the total mass of dried E-CNCs and the weighing bottle (mg), m₂ is the mass of the weighing bottle (mg), m₃ is the mass of cellulose pulp (mg), V₁ is the total volume of as-manufactured E-CNC suspension (mL), and V₂ is the volume of E-CNCs to be dried (mL).

The degree of substitution of the E-CNCs was determined by elemental analysis and chemical titration as described previously.^11,12

2.4 Field emission scanning electron microscope (FESEM) examination

The surface morphology of the samples was characterized with an XL30 ESEM-FEG model FESEM (FEI Co., Ltd., USA) with an accelerating voltage of 10 kV. All the samples were sputtered and coated with gold before observation.

2.5 Transmission electron microscope (TEM) analysis

The microstructure and size of the manufactured esterified cellulose nanocrystals were observed with a transmission electron microscope (TEM). Small drops of the E-CNC solution were deposited on a copper grid coated with a carbon support film for 1 min, and then the excessive water was removed with filter paper, followed by staining with phosphotungstic acid solution (2 wt%). The grid was dried at room temperature for 12 h, and then analyzed with a JEOL JEM-1010 TEM (Japan Electronics Co., Ltd., Japan) operated at an accelerated voltage of 100 kV.

2.6 Fourier-transform infrared (FTIR) analysis

The chemical structure changes between cellulose pulp and the manufactured esterified cellulose nanocrystals were analyzed by infrared spectroscopy. FTIR spectra of the samples were obtained with a Nicolet 380 FTIR spectrometer (Thermo Electron Instruments Co., Ltd., USA) in the range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. Prior to analysis, each sample was first ground into powder, blended with KBr and then pressed into thin pellets.

2.7 X-ray photoelectron spectroscopy (XPS) analysis

The chemical compositions of the surface of the esterified cellulose nanocrystals were characterized by X-ray photoelectron spectroscopy. X-ray photoelectron spectrograms were measured using an ESCALAB 250 (Thermo Scientific Instruments Co., Ltd., USA) X-ray Photoelectron Spectrometer equipped with a monochromated Al Kα X-ray source (1486.6 eV) and operated at a voltage of 15 kV under a current of 10 mA. Elemental surface compositions were determined from low-resolution survey measurements (100 eV pass energy and 1 eV step), and carbon surface chemistry was probed with high-resolution regional scans (30 eV pass energy and 0.05 eV step). Surface elemental concentrations and O/C ratios were calculated from the survey spectra. The carbon C1s high-resolution spectra were curve fitted using parameters defined for cellulosic materials and all binding energies were referenced to the aliphatic carbon component of the C1s signal at 285.0 eV.

2.8 Solid-state ¹³C NMR spectroscopy

The NMR experiments were performed with a Bruker Avance III 500 spectrometer (Bruker Biospin AG, Fallanden, Switzerland) operated at a ¹³C frequency of 125.73 MHz, a magic angle spinning (MAS) rate of 5 kHz and a contact time of 2 ms. All spectra were collected over 1.5 h (1024 scans).

2.9 X-ray diffraction (XRD) analysis

The crystallinity index (CrI) and crystalline structure changes of the samples were investigated by X-ray diffraction (XRD) analysis. The X-ray diffraction (XRD) measurement was performed by means of an X’Pert Pro MPD X-ray diffractometer (Philips-FEI, Netherlands) with Cu Kα radiation. The applied current and accelerating voltage were 30 mA and 40 kV, respectively. Diffractograms were collected in the range of 2θ = 6–60° at a scanning rate of 0.1° s⁻¹. The crystallinity index (CrI) was calculated according to eqn (2):

CrI = (I₀₀₂ − I_am)/I₀₀₂, (2)

where I₀₀₂ is the overall intensity of the peak at 2θ ≈ 22°, representing the crystalline part of the material, and I_am is the intensity of the baseline at 2θ ≈ 18°, representing the amorphous part of the material.¹³

2.10 Thermogravimetric analysis

Thermogravimetric analysis was conducted under a nitrogen atmosphere using a thermal analyzer (NETZSCH STA449F3, Germany). The samples were heated within a temperature interval of 25–550 °C at a constant heating rate of 10 °C min⁻¹ and a flow rate of 25 mL min⁻¹.

3. Results and discussion

3.1 Effects of reaction conditions on the yield and DS of E-CNCs

With the effect of mechanochemistry esterified cellulose nanocrystals were obtained, in which the nanocrystallization and esterification took place simultaneously. In a previous experiment we investigated the rotational speed on the yield and characteristics of E-CNCs, and found its effect was not significant. The effects of ball milling time, reaction temperature and ultrasonication time on the yield and degree of substitution (DS) were investigated here. Fig. 2a shows the effect of ball milling time on the yield and DS values, that is, at reaction temperature of 120 °C, with a reaction time of 5 h and an ultrasonication time of 3 h, the yield of E-CNC product increases from 17.45% to 43.06% with the ball milling time increasing from 1 h to 2 h. This result arises from the transfer of much more imposed stress produced during ball milling, while the action of ball milling causes the cleavage of intermolecular and intramolecular hydrogen bonds in the cellulose chains, which leads to the exposure of more free hydroxyl groups.¹⁴ In addition, continuous particle refinement and efficient mixing of acetic acid and cellulose pulp result from ball milling, which promotes sequential esterification. This yield decreases to 34.37% when the ball milling time is extended to 3 h, which may be attributed to the excessive destruction of the crystalline regions during ball milling. The graph shows that the DS values increase as the ball milling time increases to 2 h, and then drops to 0.2 as the ball milling time is extended to 3 h. This phenomenon can be explained by the initial acetylation of the disordered accessible regions of the cellulose, followed by esterification at interior regions of the cellulose crystals. As more hydroxyl groups are exposed, only some of them can be acetylated under the experimental conditions.

Fig. 2 Effect of (a) ball milling time, (b) reaction temperature and (c) ultrasonication time on the yield and degree of substitution (DS) of E-CNCs.

The effects of reaction temperature on the yield and DS values of E-CNC product are shown in Fig. 2b. With other factors constant, the yield is up to 44.82% at 120 °C and decreases to 28.84% at 140 °C, which may be explained by the fact that the reactivity is enhanced with the increase of temperature. However, at higher temperature (>120 °C), the hydrolysis of the β-glucosidic bonds of cellulose takes place, followed by excessive disintegration of cellulose pulp.

Fig. 2c shows the effect of ultrasonication time on the yield and DS values of E-CNC product. With the ultrasonication time varying from 2 h to 3 h and other parameters constant, the yield of E-CNC product increases from 17.81% to 43.21%. This is probably due to the fact that the ultrasonication treatment allows the formation, growth and collapse of cavitation bubbles in aqueous solution.¹⁵ The resulting ultrasound energy can be transferred to cellulose chains through the cavitation process, and breaks down the interaction forces between cellulose microfibrils, facilitating the disintegration of amorphous regions of cellulose, allowing the reagent to enter the interior cellulose fibers.¹⁶

3.2 Morphological analysis

The FE-SEM images of cellulose pulp and E-CNC powders are shown in Fig. 3. A curled and flat shape is observed in the SEM image of cellulose pulp (Fig. 3a), and the surface of the cellulose pulp is separated into individual micro-sized fibers. In fact, these micro-sized cellulose fibers are composed of strong hydrogen bonding nanocrystals.¹⁷ Under the effects of the mechanochemical process, individual nanocrystals could be obtained. It can be seen that the surfaces of the cellulose pulp are smooth and clean while the E-CNCs are rougher (Fig. 3b), indicating that acetylation has affected the structure of fibers. As is shown in the TEM micrographs of the E-CNCs (Fig. 3c and d), short rod-like E-CNCs are obtained with estimated widths ranging from 20–40 nm and lengths distributed between 130 and 230 nm, which is similar to other studies.¹⁸ These nanocrystals display a classical web-like network structure and occur as very long entangled cellulosic filaments, which could provide higher reinforcing capability for composite applications.¹⁹ Similar results were found in other literature, which reported that the strong H-bonding among nanocrystals led to the formation of self-assembled networks.²⁰ In addition, the lower degree of agglomeration of nanocrystal bundles may be due to the replacement of hydroxyl groups in cellulose by acetyl groups, causing the cellulose to be more hydrophobic and thus resulting in better dispersion.

Fig. 3 SEM and TEM images of cellulose pulp and E-CNCs, and the length and width distributions of the E-CNCs.

3.3 FTIR analysis

Fig. 4 shows the FTIR spectra of cellulose pulp and the E-CNC materials. In the two spectra, the absorbances at 3416, 2900, 1639, 1059, and 898 cm⁻¹ are associated with native cellulose. The bands at 3416 cm⁻¹ and 2900 cm⁻¹ are assigned to the hydrogen bond O–H stretching vibration and the C–H symmetric stretching vibration, respectively. The peak at 1639 cm⁻¹ corresponds to the O–H bending of the absorbed water. The absorption band at about 898 cm⁻¹ could be attributed to the asymmetric out-of-plane ring stretching in cellulose, which is due to the β-linkage and the amorphous part of the cellulose.^21,22 The strong absorption peak at 1059 cm⁻¹ arises from the stretching vibration of the C–O bond in cellulose, as well as the C–O bond in the hemiacetal of pyranose, which weakens in the E-CNCs because of the reduction in molecular weight.²³ As for the E-CNCs, the peak located at 1731 cm⁻¹ is attributed to the C=O stretching of the carbonyl group in the ester bonds.²⁴ The vibration peak at 1245 cm⁻¹ corresponds to the C–O stretching of the acetyl group. These two peaks confirm the acetylation of cellulose fibers. In addition, for the E-CNCs, the peak area at 3416 cm⁻¹ is lower than that of cellulose pulp, indicating partial acetylation. The peak at around 1700 cm⁻¹ is associated with the stretching vibration of the free carboxylic group in the acetic acid.²⁵ The absence of this peak indicates that the product is free of unreacted acetic acid.

Fig. 4 FTIR spectra of cellulose pulp and E-CNCs.

3.4 XPS characterization

Fig. 5 shows the XPS spectra of the cellulose pulp and E-CNC materials. The low-resolution scan model reveals that the elements present in the cellulose pulp and E-CNCs are C and O, whose peaks occur at about 284.5 eV and 533 eV, with no other elements existing in the XPS spectra. Based on the total area of peaks and the respective photoemission cross-sections, the relative distribution of the composition of O and C and the oxygen/carbon (O/C) ratio for cellulose pulp and E-CNCs were determined. For the cellulose pulp, the composition of O and C is 41.85% and 58.15%, respectively. In contrast, the composition of O and C in the E-CNCs is 39.57% and 60.43%, respectively. The O/C ratios are 0.72 for cellulose pulp and 0.65 for the E-CNCs. The decrease in O/C ratio for the E-CNCs is probably due to the acetylation of cellulose. The high-resolution scan of the C1s regions of the cellulose pulp and E-CNCs are deconvoluted into four peaks that are expressed as C1, C2, C3 and C4, which provide the relative areas of the C1 (C–H, C–C), C2 (C–O), C3 (O–C–O or C=O) and C4 (O–C=O) moieties. As for C4, it is omitted here as it is low in content. The variation of peak area contributions of the C1, C2 and C3 components in the cellulose pulp and E-CNCs show that the C1 and C2 components are the major constituents of the C1s peak in bamboo pulp. The C1 contributions increased from 19.21% to 23.80%, while the contributions of C2 decreased from 62.34% to 52.07%. This indicates the increase of C–C or C–H components and the decrease of hydroxyl groups in cellulose. The decrease in the O/C ratios and the corresponding increase in the contributions of C1 following the esterification reactions clearly show the occurrence of the expected surface modification of the cellulose fibers.

Fig. 5 XPS survey spectra of (a) cellulose pulp and (b) E-CNCs.

3.5 Solid-state ¹³C NMR spectroscopy

Fig. 6 shows the CP-MAS ¹³C NMR spectra of unmodified cellulose pulp and E-CNCs. The ¹³C NMR spectra of all the cellulose samples present the characteristic signals of cellulose I. The strong signals in the region between 60 and 120 ppm are assigned mainly to the different carbons of cellulose. The two resonances at 62 and 64.5 ppm are assigned to the disordered and crystalline regions of the C6 carbons of cellulose, and the signals between 70 and 75 ppm are assigned to the C2, C3 and C5 carbons of the glucopyranose rings in the crystalline regions.²⁶ The signals at 83.5 and 88.5 ppm are attributed to the disordered and crystalline regions of C4 carbons, respectively. The resonance at 104.5 ppm resulting from the C1 carbons of the cellulose structure is ascribed to the crystalline regions of cellulose.²⁷ The peaks at 88.5 and 64.5 ppm ascribed to the crystalline regions remain the same, which indicates that only a very small part of the crystalline region is be altered by the esterification.²⁸ This hypothesis can be supported by the XRD analysis. After esterification, the characteristic signals of the grafted moieties occur at 174 and 21 ppm, corresponding to the carbons of the carbonyl and methyl groups, respectively. The crystalline structure of cellulose samples can be also investigated by NMR spectroscopy. The crystallinity index can be obtained by evaluating the C4 signal of cellulose, however, the intensity of the C4 signal changes little, confirming that the crystalline region of cellulose is not altered.

Fig. 6 CP-MAS ¹³C NMR spectra of (a) cellulose pulp and (b) E-CNCs.

3.6 X-ray diffraction (XRD) analysis

To gain further insight into the crystalline structure changes caused by the mechanochemical process, X-ray diffraction (XRD) analysis was conducted. XRD patterns of all the cellulose samples and the calculated crystalline index are shown in Fig. 7. All the samples have similar diffraction patterns with four diffraction peaks at 2θ = 15.8°, 17.3°, 23.5° and 35.4°, corresponding to the (11̄0), (110), (200) and (004) crystallographic planes, in agreement with the characteristic diffraction peaks of cellulose Iβ,²⁹ indicating that the crystalline type of cellulose is not altered during the ball milling and ultrasonication treatments in this study. In order to investigate the evolution of the crystalline structure of cellulose during the ball milling process, XRD spectra of BM (1 h), BM (2 h) and BM (3 h) were measured and the diffractograms are shown in Fig. 7a. During the ball milling process, the peak intensity at 23.5° increases, confirming the cleavage of glycosidic linkages and the destruction of the hydrogen bond network in cellulose. Compared to cellulose pulp, the crystallinity index of BM (2 h) increases from 63.5% to 71.3%. Extending the ball milling time from 2 h to 3 h results in the decrease of the crystallinity index from 71.3% to 67.3%, this is probably due to the excessive destruction of crystalline cellulose.

Fig. 7 XRD patterns of untreated and mechanochemical treated cellulose samples. (a) XRD spectra of cellulose pulp and ball milled samples for 1 h, 2 h and 3 h, labeled as BM (1 h), BM (2 h) and BM (3 h), respectively. (b) XRD spectra of cellulose pulp, BM (2 h) and E-CNCs.

As shown in Fig. 7b, the peak intensity of the (200) plane increases after mechanochemical treatment. The calculated crystallinity index (CrI) of the cellulose pulp, BM (2 h) and the E-CNCs are 63.5%, 71.3% and 77.2%, respectively. The increase in degree of crystallinity for the E-CNCs compared to that of the cellulose pulp can be explained by the degradation of amorphous regions and disordered regions of cellulose during mechanochemical treatment. Moreover, the cellulose samples after esterification still maintain a high crystallinity index, confirming that the modification occurred essentially at the surface and possibly in the amorphous regions of the cellulose.³⁰ These results are in good agreement with the FTIR results. The crystallinity of cellulose is known as one of the main factors determining its mechanical and thermal properties. Higher crystallinity in E-CNCs is associated with higher tensile strength and thermal stability, which is expected to be beneficial for producing high-strength composite materials.

3.7 Thermogravimetric analysis (TGA)

Representative TG and DTG curves of cellulose pulp and the E-CNCs are shown in Fig. 8. The cellulose pulp shows an initial small weight loss between 25 °C and 120 °C associated with the evaporation of water molecules contained in the sample, which is not obvious for the E-CNCs. This indicates that the acetylated products are more hydrophobic than native cellulose. For the cellulose pulp, the initial thermal decomposition occurs at 286 °C, followed by a drastic reduction in weight between 286 °C and 350 °C. In contrast, the significant weight loss for the E-CNCs occurs in the range of 315–367 °C, followed by a slow weight loss up to 450 °C. Moreover, the DTG curves (Fig. 8b) show that the thermal decomposition peaks of the maximum weight loss appear at 326 °C for the cellulose pulp and 350 °C for the E-CNCs, which is due to the thermal decomposition of cellulose. All of the above results indicate that the thermal stability of the E-CNCs is higher than that of cellulose pulp. The thermal stability of cellulose is affected by crystalline order, which increases after the nanocrystallization and acetylation.³¹ This explains why the thermal stability of the E-CNCs is higher than that of cellulose pulp. The high-temperature properties of the E-CNCs may broaden the fields of application of cellulose fibers, especially at temperatures higher than 200 °C for biocomposite processing.

Fig. 8 (a) TG-curves and (b) DTG curves of cellulose pulp and E-CNCs.

4. Conclusions

Esterified cellulose nanocrystals (E-CNCs) were manufactured via one-pot tandem reactions with the catalysis of DMAP. Ball milling plays an important role in mechanochemical activation during the tandem reaction process. The mechanical shearing and friction forces generated break down the interaction force between cellulose microfibrils, leading to the nanocrystallization and acetylation of cellulose simultaneously. The resulting E-CNC particles are of 130–230 nm in length and 20–40 nm in width. In addition, the E-CNCs have a high crystallinity index (77.2%), indicating the maintenance of the crystalline region. Therefore, this study provides an efficient approach and mild operating conditions to prepare functionalized cellulose nanocrystals.

Acknowledgements

We appreciate the generous financial support of the Scientific Research Foundation of Graduate School of Fujian Agriculture and Forestry University (Grant no. 1122yb018), the Project of Advanced Forestry Science and Technology (Grant no. 2014-4-30) and the National Natural Science Foundation of China (Grant no. 31170520, 31370560).

References

1. A. Müller, M. Zink, N. Hessler, F. Wesarg, F. A. Müller, D. Kralisch and D. Fischer, RSC Adv., 2014, 4, 57173–57184.
2. L. C. Tomé, R. J. Pinto, E. Trovatti, C. S. Freire, A. J. Silvestre, C. P. Neto and A. Gandini, Green Chem., 2011, 13, 419–427.
3. S. Ifuku, M. Nogi, K. Abe, K. Handa, F. Nakatsubo and H. Yano, Biomacromolecules, 2007, 8, 1973–1978.
4. M. Pešić, C. López, J. López-Santín and G. Álvaro, Appl. Microbiol. Biotechnol., 2013, 97, 7173–7183.
5. A. Bruggink, R. Schoevaart and T. Kieboom, Org. Process Res. Dev., 2003, 7, 622–640.
6. K. Ko, K. Nakano, S. Watanabe, Y. Ichikawa and H. Kotsuki, Tetrahedron Lett., 2009, 50, 4025–4029.
7. H. Hagiwara, H. Inoguchi, M. Fukushima, T. Hoshi and T. Suzuki, Tetrahedron Lett., 2006, 47, 5371–5373.
8. A. Sakakura, K. Kawajiri, T. Ohkubo, Y. Kosugi and K. Ishihara, J. Am. Chem. Soc., 2007, 129, 14775–14779.
9. Z. Liao, Z. Huang, H. Hu, Y. Zhang and Y. Tan, Bioresour. Technol., 2011, 102, 7953–7958.
10. A. Bruckmann, A. Krebs and C. Bolm, Green Chem., 2008, 10, 1131–1141.
11. C. Freire, A. Silvestre, C. P. Neto and R. Rocha, Cellulose, 2005, 12, 449–458.
12. D.-Y. Kim, Y. Nishiyama and S. Kuga, Cellulose, 2002, 9, 361–367.
13. W. Y. Hamad and T. Q. Hu, Can. J. Chem. Eng., 2010, 88, 392–402.
14. P. Huang, M. Wu, S. Kuga and Y. Huang, Cellulose, 2013, 20, 2175–2178.
15. Q. Lu, W. Lin, L. Tang, S. Wang, X. Chen and B. Huang, J. Mater. Sci., 2015, 50, 611–619.
16. P. B. Filson and B. E. Dawson-Andoh, Bioresour. Technol., 2009, 100, 2259–2264.
17. K. Abe and H. Yano, Cellulose, 2010, 17, 271–277.
18. M. M. Haafiz, A. Hassan, Z. Zakaria and I. Inuwa, Carbohydr. Polym., 2014, 103, 119–125.
19. A. Bhatnagar and M. Sain, J. Reinf. Plast. Compos., 2005, 24, 1259–1268.
20. R. M. Sheltami, I. Abdullah, I. Ahmad, A. Dufresne and H. Kargarzadeh, Carbohydr. Polym., 2012, 88, 772–779.
21. A. Alemdar and M. Sain, Bioresour. Technol., 2008, 99, 1664–1671.
22. J. P. Reddy and J.-W. Rhim, Carbohydr. Polym., 2014, 110, 480–488.
23. X. Sun, F. Xu, R. Sun, P. Fowler and M. Baird, Carbohydr. Res., 2005, 340, 97–106.
24. V. Tserki, N. Zafeiropoulos, F. Simon and C. Panayiotou, Composites, Part A, 2005, 36, 1110–1118.
25. N. Zafeiropoulos, G. Dijon and C. Baillie, Composites, Part A, 2007, 38, 621–628.
26. P. Tingaut, T. Zimmermann and F. Lopez-Suevos, Biomacromolecules, 2010, 11, 454–464.
27. X. Wang, H. Li, Y. Cao and Q. Tang, Bioresour. Technol., 2011, 102, 7959–7965.
28. K. Zhang, E. Brendler, A. Geissler and S. Fischer, Polymer, 2011, 52, 26–32.
29. A. D. French, Cellulose, 2014, 21, 885–896.
30. L. Tang, B. Huang, Q. Lu, S. Wang, W. Ou, W. Lin and X. Chen, Bioresour. Technol., 2013, 127, 100–105.
31. J. E. Sealey, G. Samaranayake, J. G. Todd and W. G. Glasser, J. Polym. Sci., Part B: Polym. Phys., 1996, 34, 1613–1620.

---