Coagulation of Cellulose: from Ionic-Liquid Solution to Cellulose Nanostructure
A linear chain of glucose monomers, cellulose, provides the structural reinforcement of the cell walls of plants and constitutes almost half of their dry mass. Wood and other plant-based raw materials are processed on a large industrial scale to isolate the cellulose, which is then dissolved. The resulting solutions can be shaped into films or fibers and solidified as such by immersion in a nonsolvent. The properties of the solidified cellulose products can, however, vary and are frequently not quite satisfactory. In the solidification process, cellulose forms one phase and the nonsolvent and solvent form a second phase, which is later removed through washing and drying. However, these phase separations of ternary mixtures are more complicated than the sentence above indicates. In fact, the details left out decide the properties of those variable materials.
This thesis reports on the interdependencies between several parameters and aspects that are critical to cellulose phase separations: compound properties, phase equilibria for the ternary mixtures, the diffusion processes, and the nanostructures formed. Several new experimental methods were developed to measure the critical amounts of nonsolvent that induce coagulation, the mass transport of solvent and nonsolvent, and the rates of coagulation. The cellulose solutions of an ionic liquid, 1ethyl-3methyl-imidazolium acetate, [C2mim][OAc], with varied amounts of a cosolvent, DMSO, were coagulated in water, ethanol (EtOH), or 2-propanol (2PrOH). It was found that 2PrOH is, expressed in molar ratio, the strongest nonsolvent (> EtOH > water). However, the diffusive rates, D, and coagulation rates were in the opposite order (water > EtOH > 2PrOH). The observed differences between nonsolvent compounds were much larger for D[C2mim][OAc] than for DNonSolvent , for the rates of coagulation or for DDMSO, particularly with high cellulose concentration.
More differences between water and alcohol as the nonsolvent were observed in the cellulose structures formed. Coagulation in water produced relatively well-ordered crystalline structures, whereas coagulation in alcohol did not. The differences between water and alcohol as the nonsolvent can be explained by different modes of phase separation and differences in nonsolvent interactions with [C2mim][OAc] and cellulose. To show the reader how we arrived at those conclusions, which have not been found in previous literature in the cellulose field, a substantial background regarding the properties and interactions of the compounds is supplied.
Networks of cellulose nanofibrils were formed in all the nonsolvents tested, which explained the generally high diffusivities observed and the minor effect of cellulose on diffusion. It appeared that diffusion through the cellulose nanofibril network is similar to diffusion in a mixture of [C2mim][OAc] and nonsolvent only, which was confirmed with a simplistic computer model.