Separation of Biomass Components by Membrane Filtration - Process Development for Hemicellulose Recovery

Sammanfattning: One of the major challenges facing the world today is the reduction of our dependency on fossil resources as their exploitation has severe negative effects on the environment. One way of realizing this is to utilise the components of lignocellulosic biomass to a greater extent as feedstock for various industrial products. However, such utilisation requires the extraction of the biopolymers from their native structure, followed by efficient isolation from the complex extraction solution. The aim of this work was to develop processes that can be used on a large scale for the separation and recovery of biomass components from various kinds of industrial process streams. In the first part of this work, a membrane-based separation process was developed that allows the fractionation of biomass components in two different process waters originating from mechanical pulp production. Filtration and membrane filtration were optimised in order to obtain concentrated fractions of wood fibre residues, suspended/colloidal matter, hemicelluloses and lignin. The most valuable fraction obtained from this process was the ultrafiltration retentate, containing up to 64 g/l of hemicelluloses at a purity of about 60%. Furthermore, the process generated a purified water stream that could be reused in the pulp mill. Using enzymatic treatment, it was possible to cross-link hemicelluloses in pulp mill process water, thereby increasing their molecular mass. Subsequent ultrafiltration was shown to partly separate hemicelluloses with high molecular mass from those with low molecular mass. This approach could be of interest in cases where large hemicelluloses with a narrow size distribution are needed for the manufacture of bio-based products. The major objective in the second part of the work was to recover hemicelluloses from a high-viscosity solution obtained after alkali extraction from wheat bran. Hemicelluloses were separated from the extraction chemical (sodium hydroxide) using a ceramic ultrafiltration membrane with a cut-off of 10 kDa. Only a moderate flux was achieved (up to 75 l/m2h) by optimising the operating conditions. However, flux could be increased substantially (to 225 l/m2h) by the introduction of dead-end filtration prior to ultrafiltration. It was also shown that the extraction chemical could be recovered from the ultrafiltration permeate by nanofiltration. An extremely high flux and good removal of organic impurities (lignin, sugars) were achieved with the polymeric nanofiltration membranes NP010 and MPF-36. Preliminary cost estimates revealed that the investment in a nanofiltration plant would be paid back after only 1150 h of operation.