ABSTRACT
The beginning of the last three decades of studies on the structure of cellulose was marked by the reintroduction of unit cell models based on parallel alignment of the cellulose molecular chains [2,3], not unlike those abandoned by Meyer and Misch [4] in the 1930s, but also incorporating bending of the glycosidic linkage to allow the intramolecular hydrogen bond, as suggested by Hermans [5]. The new models were not consistent with each other, however, apart from the fact that both were based on parallel alignment of the cellulose chains. As French [6] pointed out, they were also not strongly preferred over an antiparallel structure. In the analysis by French [6], it was recognized that the source of the inconsistency was not so much that the different laboratories were using different computational approaches as it was that the different diffractometric data sets were gathered from different samples and represented different intensities for the same reflections. All of these studies were undertaken before the variability of the crystalline forms of native celluloses was revealed through the high-resolution solid-state 13C NMR investigations.
