Chromatographic separation of wood constituents
Doctoral thesis, 2014
In a modern kraft pulp mill only about 45 % of the incoming wood is present in the pulp and it is mainly composed of cellulose. The remainder of the wood, which is dissolved in order to liberate the cellulose fibres, e.g., hemicellulose and lignin, ends up in the mill´s recovery boiler to recover the latent energy. The biorefinery concept aims to utilise a larger proportion of the biomass feedstock by converting, for example, hemicelluloses and lignin into new materials, chemicals or fuel.
The research in this thesis investigates the potential and limitations of using chromatographic separation to purify fractions of biopolymers extracted from wood. In contrast to other means of separating wood constituents, chromatography is a highly selective and non-destructive separation process that utilises the species differences in chemical structure. A major drawback associated with this high selectivity is the slow rate of separation, which, in turn, leads to high separation costs. The challenge of using chromatography to separate biopolymers is to develop a method with high specificity while maintaining low costs of separation. To achieve this, a suitable chromatographic system must be developed, static process parameters must be optimised and mass transfer resistance must be minimised.
The findings of this thesis show that wood biopolymers can be separated according to relative contents of aromatic groups using hydrophobic adsorption. It was also found that the solubility of the biopolymers is a limiting factor in the rate of production. Furthermore, calculations indicated that the dominant mass transfer resistance in the system is the intra-particle diffusivity, and therefore the sorbent particle radius should be low (15-20 µm), for an efficient system. Separating larger polymers will also have a negative influence on efficiency, due to dispersive, diffusive, steric and solubility effects. An analysis of the economics involved indicates that the process is not economically viable for the purpose of producing new materials. However, a model system was created that can be used in pre-studies to develop a flexible semi-preparative system for use in fundamental science to better characterise wood constituents.