Infrared spectra of cellulose and related polysaccharides

Abstract

Abstract A comparison has been made between the infrared spectra of native cellulose and those of various closely related materials differing in each instance from cellulose I in one salient feature. These materials have been compared individually with groups of others, each differing from cellulose in a further salient feature. Considered in conjunction with the spectra of a few formally more distant relatives such as tetrahydropyran and the simple alcohols, these empirical comparisons can contribute to an understanding of the influence on the spectrum of hydrogen bonding, hydroxyl groups (through HD exchange and o ‐acylsubstitution), the type of anomeric and positional linkage, attachment at a specific carbon atom, the glycosidic link, and degree of polymerization. The comparative spectroscopy of cellulose, laminarin, amylose, and xylan permits correlation between these glucosans for about a dozen of the major bands or groups of bands in the spectra of the amorphous materials between ca. 3 and 13μ. An examination of the spectra of various crystalline modifications of cellulose lends support to various conclusions drawn by previous workers regarding the relationship between these forms. The effects of deuterating native and regenerated cellulose have been taken into account in considering band assignments. The intensity of the strong band near 8.6 μ (1160 cm. −1 ) is reduced on deuteration, and it is questioned whether this is consistent with its assignment to stretching in the glycosidic link (on the basis of its strong parallel polarization, etc.). The response to deuteration suggests that the band near 9.0 μ (1120 cm. −1 ) may be an association band of the type observed in primary and secondary alcohols and that the residual band near 8.9 μ (1120 cm. −1 ) may be identified with the antisymmetric ring‐stretching vibration. The persistence of the strong bands between 9 and 10 μ (1100–1000 cm. −1 ) after deuteration, acetylation and, in the case of sugars, freeze drying, supports the conventional assignment to CO stretching, perhaps coupled to CC frequencies. The band near 10.1 μ (990 cm. −1 ) is eliminated on deuteration or hetero‐geneous acetylation of cellulose. These and other facts suggest that it is connected with a vibration which is restricted in the crystallites. The band a t 11.2 μ (893 cm. −1 ) in cellulose and other β‐linked glucosans is clearly connected with the C (1) group vibration. In much of this work the samples were examined as dispersions in pressed potassium chloride pellets. In the amorphous state the spectra of the smaller molecules resemble those of the corresponding polymers. It was also observed that freeze‐dried samples, dispersed in potassium chloride, showed a slow reversion to the sharp spectrum characteristic of the crystal.