Highly crystalline cellulose samples, which also have a high purity of crystal form, were used or prepared as models of cellulose polymorphs, cellulose I, III,, IV,, II, IIIn, and IVn, for solid-state 13C NMR and X-ray diffraction analyses. Differences between cellulose I, II, and III in the 13C NMR spectra appear at the chemical shifts of C6 in the anhydroglucose units; they have signals at 65.5-66.2, 63.5-64.1, and 62.1-62.8 ppm, respectively. Cellulose IV has a doublet signal for C4 at 83.6-84.6 ppm and a signal for C6 at 63.3-63.8 ppm, which is identical with that of cellulose II. Thus the mechanisms for the formation of cellulose polymorphs are primarily ascribed to some conformational and/or environmental transformations of C4 and C6. So-called cellulose IV, is likely to be a mixture of cellulose I and real cellulose IV,. The differences between cellulose III, and III,, were not detected clearly at the chemical shifts of C4 and C6 but at the chemical shift of Cl and the signal pattern of C2, C3, and C5.

SYNOPSISChemical structures of cellulose and chitosan dissolved in trifluoroacetic acid (TFA) and those of cellulose and chitosan films cast from their TFA solutions were studied by 13C-NMR and infrared ( IR) spectroscopy. Cellulose is trifluoroacetylated selectively at the C6-hydroxyl groups in the TFA solution, and chitosan is dissolved in TFA by forming amine salts with TFA a t the CP-amine groups. IR analyses of cellulose films cast from its TFA-acetic acid solutions showed that partly trifluoroacetylated cellulose in the solution state turns to partly acetylated cellulose in the solid state during evaporation of the solvents in air by the ester interchange. Chitosan films cast from its TFA-acetic acid solutions still have the amine salts with TFA. These acetyl groups in cellulose films and TFA in chitosan films are removable by soaking the films in 1N NaOH at room temperature for 1 day.

SYNOPSISInteractions between cellulose and chitosan molecules in cellulose-chitosan blend films, prepared using trifluoroacetic acid as a cosolvent for the two polysaccharides, were studied by X-ray diffraction and Raman analyses and by measurements of mechanical properties of the blend films. Crystallinity of cellulose in the blend films decreased with an increase in chitosan content. The blend films had tensile strengths of 45-100 MPa and Young's moduli of 2-7.5 GPa in dry states. These values had the maximum around 30% chitosan content in the blend films. These results suggested the presence of interactions between cellulose, chitosan, and water molecules in the films. However, Raman analysis suggested that cellulose and chitosan molecules in the blend films seemed to have the same secondary structures as those in 100% cellulose and 100% chitosan films, respectively. Thus, these results indicate the presence of interactions in the interfacial region between small domains of cellulose and chitosan. The presence of chitosan molecules may lead to decrease in the domain size of cellulose, and to increase in the interfacial region between cellulose and chitosan domains.

A major unsolved problem in cellulose structure is the polarity of adjacent chains of cellulose in the microfibrils, namely, the "parallel or antiparallel" problem. This issue is closely related to the possibility of a folded-chain structure and to the mechanism of cellulose biosynthesis. The prevailing view, based on x-ray (and electron) diffraction,I4 is that native celluloses (Cellulose I) are arranged in parallel and that regenerated celluloses (Cellulose 11) are antiparallel. However, the limited size of single cellulose crystals hinders obtaining convincing results, and conflicting views are not yet ruled out? In this study, we attempted to obtain direct evidence of the molecular arrangement of cellulose in microfibrils, using a novel method with electron microscopy. When properly prepared fragments of microfibrils are selectively stained with a heavy metal at the reducing end groups, either of the two modes of staining could be observed, according to the molecular arrangement, as in Figure 1.The cell wall of Vulonia (a kind of green algae) is a convenient material for structural studies of cellulose because of its high crystallinity and exceptional microfibril width. However, it is not degraded to microcrystallites merely by heating with acids, because of its defect-free structure, so the microcrystallites were prepared as follows. Vesicles of Vulonia macrophysa purified by successive alkali and acid treatments2 were disintegrated to microfibrils by a laboratory homogenizer; at the same time the microfibrils were damaged, showing n a n y kinks and bucklings (Fig. 2). A hot-acid treatment (60 min in 20% sulfuric acid at 100°C) resulted in a suspension of rodlike fragments of various length, with the acid probably attacking the disordered regions preferentially. The chain ends of the cellulose exposed at the tips of these crystallites would then become ordinary reducing or nonreducing ends as a result of hydrolytic cleavage.By analogy with periodic acid-methenamine-silver staining used in histochemical identification of polysacchar;ldes, the reducing ends of the cellulose would be expected to be oxidized by a silver-ammonia reagent into carboxyls, onto which silver would be deposited. This procedure was applied to our cellulose sample; however, it failed to produce a sufficient density of silver stains.The procedure was modified thus: the crystallites were treated first with 50 m M sodium chlorite (pH adjusted to 3.5 with acetic acid) for 20 h at room ' Present address: Oji Paper Co.

SYNOPSISA thin membrane of bacterial cellulose (BC) obtained from Acetobacter culture was tested for its performance as a dialysis membrane in aqueous systems. The BC membrane showed superior mechanical strength to that of a dialysis-grade regenerated cellulose membrane, allowing the use of a thinner membrane than the latter. As a result, the BC membrane gave higher permeation rates for poly(ethy1ene glycols) as probe solutes. The cutoff molecular weight of the original BC membrane, significantly greater than that of regenerated cellulose, could be modified by concentrated alkali treatments of the membrane. The nature of the change at the ultrastructural level caused by the alkali treatments was studied by X-ray diffraction and scanning electron microscopy. 0 1993 John Wiley & Sons, Inc.I NTRO D U CT 1 0 N Acetobacter xylinum, a Gram-negative bacterium, produces cellulose extracellularly. This cellulose is formed as gel-like mass (pellicle) at the surface of the medium and can be purified by proper chemical treatments. This material has high crystallinity and large surface area and has been attracting attention as a new form of cellulosic material. The proposed application includes an acoustic vibrator taking advantage of its high elastic modulus's2 and an insoluble thicker/binder for foods and sheetlike mater i a l~.~ When a purified pellicle is dried on a flat substrate, a thin translucent cellulose membrane is formed. This membrane is expected to have unique properties because it consists of fine and continuous crystalline microfibrils, not like paper sheets or regenerated cellulose films. One possible application is molecular filtration such as dialysis or ultrafiltration. It has been proposed to use the bacterial cellulose as a dialysis membrane in nonaqueous syst e m~.~ On the other hand, regenerated cellulose membranes have been widely used as a dialysis membrane in aqueous systems, where chemical stability and low toxicity of cellulose are preferable properties, especially in applications for labile biological systems. This study aimed at elucidating the basic characteristics of bacterial cellulose membrane as molecular separation medium in aqueous conditions, together with modifying the structure of the membrane by chemical treatments for controlling its molecular permeation characteristics. The tests were conducted in the dialyzing mode, i.e., without pressurizing the primary-side solution, by using a series of poly (ethylene glycols) [ poly (ethylene oxide) ] as probe solutes. EXPERIMENTAL Bacterial Cellulose (BC) MembraneThirty milliliters of sterilized Schramm-Hestrin medium4 was placed in a plastic Petri dish (inner diameter, 87 mm) and was inoculated with Acetobacter xylinum ( ATCC 23769 ) . Immediately after inoculation, the medium was gently but thoroughly mixed by swirling the dish so that the cells were distributed uniformly. The culture was statistically incubated at room temperature for 5-7 days, until the liquid medium was filled with cellulose pellicle.The harvested pellicle was rinsed with...