Biomacromolecules 2000, 1, 488-492
488
Periodate Oxidation of Crystalline Cellulose Ung-Jin Kim,*,† Shigenori Kuga,† Masahisa Wada,† Takeshi Okano,† and Tetsuo Kondo‡ Department of Biomaterials Science, Graduate School of Agricultural and Life Sciences, Yayoi 1-1-1, Bunkyo-ku, Tokyo 113-8657, Japan; and Forestry and Forest Products Research Institute (FFPRI), P.O. Box 16, Tsukuba Norin, Ibaraki, 305-8687 Japan Received April 17, 2000; Revised Manuscript Received June 16, 2000
Conversion of 1,2-dihydroxyl groups to dialdehyde by periodate oxidation is a useful method of derivatizing polysaccharides but has not been extensively utilized in derivatization of cellulose because of complicacy due to the crystalline nature of cellulose. To understand the influence of cellulose crystallinity on this reaction, we investigated how the periodate oxidation proceeds with a highly crystalline cellulose of the marine alga Cladophora sp. The crystallinity of the oxidized cellulose, determined by X-ray diffraction, decreased according to the oxidation level. The half-height widths of equatorial diffraction peaks were nearly unchanged. The solid-state 13C NMR spectra did not show peaks corresponding to aldehyde groups, but solution 13C NMR spectra showed the presence of dicarboxylic groups after subsequent oxidation by sodium chlorite. Transmission electron microscopy showed that microfibrils of Cladophora tended to be bent and more flexible than the original sample. Gold labeling of the aldehyde groups, mediated by thiosemicarbazide derivatization, revealed a highly uneven distribution of dialdehyde groups. When treated by 50% (w/v) sulfuric acid, partially oxidized Cladophora cellulose gave many short fragments of microfibril. These features indicate that the periodate oxidation proceeds by forming dialdehyde groups in longitudinally spaced, bandlike domains. Introduction Periodate oxidation is a highly specific reaction to convert 1,2-dihydroxyl (glycol) groups to paired aldehyde groups without significant side reactions and is widely used in structural analysis of carbohydrates.1-6 When applied to 1,4glucans, this reaction cleaves the C2-C3 bond. This reaction is industrially utilized in production of oxidized starch.7,8 Application or quantitative understanding of this reaction with cellulose, however, has been hampered by complicacy arising from hemiacetal formation of aldehyde and crystalline nature of cellulose.9-11 Still, the resulting aldehyde compound (“dialdehyde” cellulose, DAC) can be further converted to carboxylic groups,3,12-14 primary alcohols,15 or imines (Schiff bases) with primary amines,16-18 which can serve as useful intermediates for cellulose-based specialty materials, such as an adsorbent of heavy metal,3,16 protein,17,19 or dye.20 We prepared in this study a series of dialdehyde cellulose with different oxidation levels and investigated the features of this reaction for highly crystalline cellulose obtained from a marine alga in comparison with a common cellulose sample from a higher plant. The sample Cladophora sp. is a green alga having cell walls composed of highly crystalline cellulose microfibrils similar to those of Valonia sp. Experimental Section Cellulose Materials. The samples used in this study were Whatman CF11 cellulose powder (Whatman International * Corresponding author. Telephone: +81-3-5841-7513. Fax: +81-35684-0299. E-mail address:
[email protected]. † Graduate School of Agricultural and Life Sciences. ‡ Forestry and Forest Products Research Institute.
Ltd.), of cotton origin, and Cladophora sp. cellulose, composed of ca. 20 nm × 20 nm wide crystalline microfibrils. The Cladophora cellulose sample was prepared by treating the cell wall material in 5% KOH and 0.3% sodium chlorite for several times.21 Its microcrystalline suspension was prepared by treating the material with 4 N HCl at 80 °C for 4 h with continuous stirring. The sample was repeatedly washed with deionized water by centrifugation at 1600 g and dialyzed to neutrality. This method is known to give a finely divided microcrystalline cellulose suspension without introducing a significant amount of surface charge as in the case of sulfuric acid treatment.22 Periodate Oxidation. For Whatman CF11 cellulose, 5 g of cellulose was mixed with 250 mL of 0.16 M sodium metaperiodate solution (1.3 mol for 1 mol of glucopyranose). For Cladophora cellulose, 100 mL of 3.8 g/L cellulose suspension was mixed with 100 mL of 0.25 M sodium metaperiodate solution (10.7 mol for 1 mol of glucopyranose). The mixtures were stirred gently at 20 °C in the dark for desired reaction times. After the excess periodate was decomposed with ethylene glycol, the products (DACs) were recovered and washed by centrifugation. Preparation of Dicarboxylic Acid Cellulose (DCC). Freeze-dried DAC-Whatman CF11 and never-dried DACCladophora were oxidized23 by 0.4 M sodium chlorite in acetate buffer at 40 °C for 8 h (CF11) or 20 h (Cladophora). The dicarboxylic acid cellulose was precipitated by pouring the solution into a large amount of ethanol, washed repeatedly by centrifugation with ethanol, and air-dried. Determination of Aldehyde Content. The DAC samples were converted to oximes by Schiff base reaction with
10.1021/bm0000337 CCC: $19.00 © 2000 American Chemical Society Published on Web 07/27/2000
Periodate Oxidation of Crystalline Cellulose
Figure 1. Time course of periodate oxidation of Whatman CF11 (filled circle) and Cladophora cellulose (filled rectangle).
hydroxylamine. The reagent (0.0125 mol) was dissolved in 100 mL of pH 4.5 acetate buffer (0.1 M) and added to 10 mL of DAC suspension containing 0.1 g (Whatman CF11) or 0.05 g of solid (Cladophora). The mixture was stirred at 20 °C for 24 h and the product was recovered by centrifugation. The elemental composition was determined for C, H, and N by atomic absorption spectroscopy. X-ray Diffraction. The X-ray diffractometry profiles of DACs were recorded for dry pellets in reflection mode (Rigaku RINT 2000, at 30 kV and 40 mA) with Ni-filtered Cu KR radiation (λ ) 0.154 18 nm). FTIR Spectroscopy. Approximately 1 mg of dry cellulose sample was pressed into a pellet with 200 mg of potassium bromide and Fourier transform infrared (FTIR) spectrum was recorded by Nicolet Magna 860 with accumulation of 64 scans and a resolution of 4 cm-1. Solid-State 13C NMR. A Chemagetics CMX-300 NMR spectrometer was used to obtain solid-state 13C NMR spectra with cross-polarization-magic-angle spinning (CP-MAS) operating at 75.57 MHz. The chemical shift was determined using hexamethylbenzene as external standard. Solution 13C NMR of Dicarboxylic Acid Cellulose. 13C NMR spectra were obtained by Bruker AC-300 from solutions in D2O (99.7%), using 3-trimethylsilyl-2,2,3,3-d4propionic acid sodium salt (δ ) 0.0 ppm) as the external standard. Transmission Electron Microscopy. Original and periodate-oxidized Cladophora cellulose were mounted on carboncoated electron microscope grids from dilute suspensions. Samples were observed with a JEM-2000EXII electron microscope operated at 200 kV. Introduction of dialdehyde groups was visualized by colloidal gold labeling as follows: the suspension of Cladophora cellulose was mixed with 1% (w/w) thiosemicarbazide in 5% (w/w) acetic acid. The mixture was kept at 60 °C for 90 min, and the solid was collected and washed by centrifugation.24 The sample was resuspended in distilled water, mixed with colloidal gold solution (Polysciences, Inc., Warrington, PA; having diameters of 15-25 nm) for 12 h in the dark, and washed with distilled water by centrifugation.
Biomacromolecules, Vol. 1, No. 3, 2000 489
Figure 2. Changes in X-ray diffraction pattern by periodate oxidation of Whatman CF11.
Figure 3. Changes in X-ray diffraction pattern by periodate oxidation of Cladophora cellulose.
Results and Discussion Figure 1 shows the time course change in the degree of oxidation (D.O.) determined from the nitrogen content of the corresponding oximes. Whatman CF11 treated with 1.3 times equivalent of sodium metaperiodate reached D.O. ) 0.80 in 190 h. In contrast, Cladophora cellulose reached only D.O. ) 0.30 even after 260 h with 10.7 times an equivalent of periodate. This behavior indicates the strong influence of cellulose crystallinity on the rate of periodate oxidation, which proceeds without dissolution of the product. The degree of crystallinity of Whatman CF11 and Cladophora, calculated from the integral intensity, was about 60% and 90%, respectively. Figure 2 shows the loss of crystallinity by periodate oxidation of Whatman CF11. The well-defined cellulose I pattern diminished with the increase in D.O. and the product became completely amorphous at D.O. ) 0.80. Figure 3 shows the corresponding change in crystallinity of Cladophora cellulose for D.O. ) 0 and 0.30. The decrease in reflection peak height was similar to the case of Whatman CF11. The loss of crystallinity is considered to result from
490
Biomacromolecules, Vol. 1, No. 3, 2000
Kim et al.
Figure 4. FTIR spectra of DAC-Whatman CF11.
opening of glucopyranose rings and destruction of their ordered packing. A notable feature for both samples is that the widths of diffraction peaks seem nearly constant in the course of oxidation and only the peak heights diminished. Interpretation of this feature will be discussed later. The changes in chemical structure by periodate oxidation were examined by FTIR and solid-state 13C NMR spectra. The DAC-Whatman CF11 and the DAC-Cladophora samples gave nearly identical spectra for the same D.O. levels. Figure 4 shows the FTIR spectra of DAC-Whatman CF11. The characteristic bands of DAC appeared in the 1740 and 880 cm-1 regions. They increased from a small shoulder to a distinct peak with increasing oxidation level. A diffuse band at 880 cm-1 can be assigned to the hemiacetal and hydrated form.1,9 The sharp peak at 1740 cm-1 is a characteristic band of carbonyl groups.1,10,11 The solid-state 13C NMR spectra showed gradual shifts of peaks from those of original cellulose I (Figure 5). The spectrum of D.O. ) 0.60 product from Whatman CF11 seems to be a superposition of the original sample and D.O. ) 0.80 product. A characteristic feature is that the lack of carbonyl signal expected at 175-180 ppm even at high oxidation degrees. This indicates that the aldehyde groups of solid DAC are hydrated or forming hemiacetals with remaining hydroxyl groups. To further confirm introduction of dialdehyde groups, the DAC was oxidized to dicarboxyl cellulose (DCC) by chlorite oxidation. The DCC-Whatman CF11 sample (D.O. ) 0.80) was readily soluble in water. The DCC-Cladophora cellulose (D.O. ) 0.30), on the other hand, was only partially soluble in water, and the residue (about 50%) was removed by filtration through 0.45 µm membrane filter. Figure 6 shows the solution 13C NMR spectrum of the soluble fraction of DCC-Cladophora. The signals of C2 and C3 of original cellulose disappeared completely and shifted to those at 177 and 179 ppm characteristic of carboxylic group. The distinct six signals show that two carboxylic groups were quantitatively introduced to the glucopyranoside unit. These results show that the solubilized part of the DCC consists of fully
Figure 5. Solid-state
13C
NMR spectra of DAC-Whatman CF11.
Figure 6. Solution 13C NMR spectra of soluble fraction of (A) DCCWhatman CF11 (D.O. ) 0.80) and (B) DCC-Cladophora (D.O. ) 0.30) in D2O.
carboxylated segments, not containing unoxidized glucopyranoside moieties. Morphological features of the DAC-Cladophora cellulose were examined by transmission electron microscopy. The original microcrystalline Cladophora sample consists of long and straight rods as seen in Figure 7A. The periodateoxidized material (D.O. ) 0.30) retained the same microfibrillar morphology but they tended to be bent and appeared more flexible than the original (Figure 7B). This
Periodate Oxidation of Crystalline Cellulose
Figure 7. Transmission electron micrographs (TEM) of (A) Cladophora cellulose and (B) DAC-Cladophora (D.O. ) 0.30). Scale bar: 500 nm.
apparently reflects the change in rigidity of microfibrils caused by cleavage of glucoside rings. Also notable is that the microfibril tips show fading out appearances in contrast to the sharp-cut ends of the starting material. This feature is considered to indicate greater reactivity of dihydroxyl groups at the tips through reduced steric hindrance. The presence of aldehyde groups on microfibrils of DACCladophora samples was visualized by labeling with colloidal gold by introduction of thiosemicarbazide as Schiff base. Figure 8 shows the microfibrils of DAC-Cladophora (D.O. ) 0.10) strongly labeled by gold particles (Figure 8B) and scarcely labeled control material (Figure 8A). The labeling in the latter probably arises from the reducing ends of original cellulose. The extent of labeling increased with the degree of oxidation, and the sample of D.O. ) 0.30 was more densely covered with gold (micrograph not shown). The cellulose microfibrils of Cladophora are similar to those of Valonia, having cross sections of approximately 20 nm × 20 nm on average. This size corresponds to ca. 1300 glucan chains, and some 10% of them should be exposed at the surface. Therefore, if the reaction proceeds from the surface uniformly, the entire surface should be oxidized at D.O. ) 0.10. Actually, however, the microfibrils was only lightly labeled at D.O. ) 0.10. This feature indicates that the oxidant’s attack on crystalline cellulose proceeds highly heterogeneously, leading to a model depicted in Figure 9. The model suggests that the oxidation of crystalline cellulose is a self-accelerating process; i.e., when a glucopyranose ring on the surface is converted to dialdehyde, the neighboring groups become more susceptible to oxidant because of the local loss of crystalline order. This model can explain the observed features of the constant half-height widths of X-ray
Biomacromolecules, Vol. 1, No. 3, 2000 491
Figure 8. TEM images of (A) Cladophora cellulose and (B) DACCladophora (D.O. ) 0.10) labeled by colloidal gold. Scale bar: 500 nm.
Figure 9. Mode of periodate oxidation of crystalline cellulose.
diffraction in the course of oxidation. Such a mode of reaction is in contrast with previously reported features of homogeneous acetylation of crystalline cellulose, for which thinning of microfibrils was observed.25 Further support for this model was obtained from TEM examination of the acid hydrolysis residue (by 50% (w/v) sulfuric acid for 30 min at 60 °C) of the DAC sample (D.O. ) 0.06). This sample contained many short fragments of microfibril with the same width as the starting material (Figure 10). Since the dialdehyde form is more susceptible to hydrolytic attack than the unmodified cellulose,26 this treatment is considered to degrade the oxidized parts preferentially. Therefore, the observed rodlike particles are the crystalline cellulose that escaped oxidation. Such short fragments have not been obtained by direct hydrolysis of
492
Biomacromolecules, Vol. 1, No. 3, 2000
Kim et al.
Acknowledgment. We thank Prof. A. Isogai of the University of Tokyo for help in solution NMR analysis, and Mrs. Y. Naito for assistance in elemental analysis. References and Notes
Figure 10. TEM images of acid hydrolysis residue of DACCladophora (D.O. ) 0.06) by 50% (w/v) H2SO4 (60 °C, 30 min).
the same material.27 The aqueous suspension of rodlike microcrystalline cellulose is known to form characteristic liquid crystalline phases and the mechanism of their formation has not been elucidated.28,29 The formation of short microcrystals by periodate oxidation-acid hydrolysis may be useful in controlling and elucidating the nature of such rodlike colloid systems. Conclusion Our results revealed several remarkable features of periodate oxidation of microfibrillar Cladophora cellulose: (i) The peak widths of X-ray diffraction of remaining cellulose did not change even when crystallinity was significantly reduced. (ii) The microfibrils with fairly high degree of oxidation was only partially labeled with colloidal gold. (iii) The acid hydrolysis residue of partially oxidized sample was composed of very short fragments having the same width as the starting material, which have never been obtained by hydrolysis of unoxidized materials. On the basis of these observations, we propose a model in which the reaction proceeds highly heterogeneously, forming isolated oxidized domains along the microfibrils. Our results emphasize the influence of crystallinity on the periodate oxidation of cellulose and the importance of the choice of starting material for attaining high degrees of conversion or controlling the distribution of dialdehyde groups.
(1) Gal’braikh, L. S.; Rogovin, Z. A. In Cellulose and Cellulose DeriVatiVes, 2nd ed.; Bikales, N. M.; Segal, L. Eds.; WileyInterscience: New York, 1971; Vol. V, Part V, p 893. (2) Nevell, T. P. In Methods in Carbohydrate Chemistry; Whistler, R. L. Eds.; Academic: New York, 1963; Vol. III, p 164. (3) Maekawa, E.; Koshijima, T. J. Appl. Polym. Sci. 1984, 29, 2289. (4) Bruneel, D.; Schacht, E. Polymer 1993, 34, 2628. (5) Schacht, E.; Bogdanov, B.; Bulcke, V.; Rooze, N. React. Funct. Polym. 1997, 33, 109. (6) Uraz, I.; Gu¨ner, A. Carbohydr. Polym. 1997, 34, 127. (7) Michell, J. H.; Purves, C. B. J. Am. Chem. Soc. 1942, 64, 585. (8) Veelaert, S.; Polling, M.; Wit, D. Starch/Sta¨ rke 1994, 46, 263. (9) Rowen, J. W.; Forziati, F. H.; Reeves, R. E. J. Am. Chem. Soc. 1951, 73, 4484. (10) Spedding, H. J. Chem. Soc. 1960, 3147. (11) Kuniak, L.; Alince, B.; Masura, V.; Alfo¨di, J. SVensk Papperstidn. 1969, 72, 205. (12) Maekawa, E.; Koshijima, T. J. Appl. Polym. Sci. 1990, 40, 1601. (13) Crescenzi, V.; Dentini. M.; Meoli, C.; Casu, B.; Naggi, A.; Torri, G. Int. J. Biol. Macromol. 1984, 6, 142. (14) Varma, A. J.; Chavan, V. B.; Rajmohanan, P. R.; Ganapathy, S. Polym. Degrad. Stab. 1997, 58, 257. (15) Casu, B.; Naggi, A.; Torri, G.; Allegra, G.; Meille, S. V.; Cosani, A.; Terbojevich, M. Macromolecules 1985, 18, 2762. (16) Koshijima, T.; Tanaka, R.; Muraki, E.; Yamada, A.; Yaku, F. Cellulose Chem. Technol. 1973, 7, 197. (17) Kobayashi, M.; Suzawa, I.; Ichishima, E. Agric. Biol. Chem. 1990, 54, 1705. (18) Maekawa, E.; Koshijima, T. J. Appl. Polym. Sci. 1991, 42, 169. (19) Gurvich, A. E.; Lechtzind, E. V. Mol. Immunol. 1982, 19, 637. (20) Lewis. D. M.; Tapley, K. N. Textile Chem. Colorist 1999, 31, 20. (21) Sugiyama, J.; Chanzy, H.; Maret, G. Macromolecules 1992, 25, 4232. (22) Araki, J.; Wada, M.; Kuga, S.; Okano, T. Colloids Surf. A 1998, 142, 75. (23) Wilson, W. K.; Padgett, A. A. TAPPI 1955, 38, 292. (24) Kuga, S.; Brown, R. M. Carbohydr. Res. 1988, 180, 345. (25) Sassi, J.-F.; Chanzy, H. Cellulose 1995, 2, 111. (26) Gladding, E. K.; Heidt, L. T.; Purves, C. B. Pap. Trade J. 1945, 121, 35. (27) Imai, T.; Sugiyama, J. Macromolecules 1998, 31, 6275. (28) Dong, X. M.; Kimura, T.; Revol, J.-.F.; Gray, D. G. Langmuir 1996, 12, 2076. (29) Araki, J.; Wada, M.; Kuga, S.; Okano, T. Langmuir 2000, 16, 2413.
BM0000337