Reconstituted nitrocellulose with significant molecular alignment

Abstract

The use of wide-angle X-ray diffraction (WAXD) to determine molecular conformation and ordering in polymers is facilitated if specimens have a high degree of molecular orientation, since this allows the scattering to be separated into interand intra-molecular components [1]. Accordingly, specimens in the form of uniaxially oriented fibres are particularly useful. Nitrocellulose (NC), however, is usually produced by nitrating either cotton linters (fibres which are ordinarily too short for use in the textile industry) or wood pulp to the required level. The material therefore typically consists of bent fibres which are no more than a couple of millimetres long, and which are practically impossible to straighten out and align. To overcome this difficulty, previous workers [2, 3] have used nitrated ramie, or specially nitrated textile grade cotton fibres. There are differences in the detail of diffraction patterns obtained from these alternative sources; they may be due to differences in the exact molecular conformations, and also due to differences in their memory for the supramolecular ordering laid down in the native cellulose by the source organisms. Certainly, the principal refractive indices of cotton fibres (1.580 and 1.534) and of ramie (1.594 and 1.532) are very different [2]; also, the swelling behaviour of cellulose on mercerization is unlike that of ramie [2] and thus reflects significant differences in supramolecular ordering. There is much to be gained, therefore, by dissolving NC and then attempting to reconstitute it in the form of aligned fibres. This would allow nitrated cotton linters or wood pulp to be used as a starting material, and the molecular ordering deduced from subsequent W A X D studies should be independent of any structural features peculiar to the original cellulose source organism. However, as far as we are aware, attempts to produce a reconstituted fibre with significant molecular alignment (as detected by WAXD) have not previously been successful. The solubility of N C in a particular solvent may vary considerably according to the polymer nitration level. The model N C solvent, both in terms of its action over a wide range of nitration levels and its prominence in the literature, would appear to be acetone. Nevertheless, it may take over a day to dissolve sufficient polymer even to make a uniform 5vo1% solution; the time taken increases with decreasing nitration level of the polymer. It was found that tetrahydrofuran (THF) acts much more rapidly on a timescale of only a few hours. Above a certain concentration, depending on the polymer nitration level (Table I), the resulting solutions have liquid crystalline textures, as is evident from the birefringence of thin specimens observed microscopically between crossed polars (Fig. 1). NC has previously been reported as showing liquid crystalline characteristics in a variety of other solvents, including acetone [4]. I f an at tempt is made to draw fibres from the anisotropic solutions in THF, it is found that the fluid has insufficient cohesivity. At high polymer concentrations ( > 40 vol %), the solutions in T H F have the mechanical behaviour of a brittle gel, and do not so readily form textures in thin specimens. Attempts to overcome the cohesivity problem by increasing the polymer content of the solutions fail because of this gel formation.