Membrane fouling is the major factor limiting the wider applicability of the membrane-based technologies in water treatment and in separation and purification processes of biorefineries, pulp and paper industry, food industry and other sectors. Endeavors to prevent and minimize fouling requires a deep understanding on the fouling mechanisms and their relative effects. In this study, Brunauer-Emmett-Teller (BET) nitrogen adsorption/desorption technique was applied to get an insight into pore-level membrane fouling phenomena occurring in ultrafiltration of wood-based streams. The fouling of commercial polysulfone and polyethersulfone membranes by black liquor, thermomechanical pulping process water and pressurized hot-water extract was investigated with BET analysis, infrared spectroscopy, contact angle analysis and pure water permeability measurements. Particular emphasis was paid to the applicability of BET for membrane fouling characterization. The formation of a fouling layer was detected as an increase in cumulative pore volumes and pore areas in the meso-pores region. Pore blocking was seen as disappearance of meso-pores and micro-pores. The results indicate that the presented approach of using BET analysis combined with IR spectroscopy can provide complementary information revealing both the structure of fouling layer and the chemical nature of foulants.

Electrodialysis (ED) is a promising emerging electrochemical membrane technology for nutrient concentration and recovery from wastewater. However associated environmental safety aspects have to be assessed before utilizing concentrated nutrient produced by ED, for instance as fertilizer. Municipal wastewaters contain various micropollutants that have the potential of being concentrated during the ED treatment processes. This study quantified the transport of pharmaceuticals during ED nutrient recovery from synthetic centrate wastewater. Specifically, it is evaluated whether pharmaceutical micropollutants are mobile, and therefore able to transport across the cation exchange membranes and concentrate into the ED concentrate product. Results demonstrate that NH₄⁺-N, PO₄^3--P and K⁺ could be concentrated up to 5 times in the concentrated ED product (3700-4000 mg/L NH₄⁺-N, 21-25 mg/L PO₄^3--P, 990-1040 mg/L K⁺). Target micropollutants, such as diclofenac, carbamazepine and furosemide were largely retained in the diluent, with less than 8% being transported across to the concentrate product (feed micropollutant concentration 10 or 100 μg/L) based on the final target pharmaceutical amounts in the ED concentrate product (μg). Some transport of micropollutants such as atenolol, metoprolol and hydrochlorothiazide was observed to the concentrate product. For instance a final concentration of 10.3, 9.4 and 8.6 μg/L on average was measured for these pollutants in the final ED concentrate product (final volume ∼1 L) in experiments with a feed water (initial volume 20 L) containing only 10 μg/L of target pharmaceuticals. Transport of pharmaceuticals across the ED membranes was concluded to be dominated mainly by the molecule hydrophobicity/hydrophilicity as well as electrostatic interactions between pharmaceutical molecules and ED membranes. Particularly excluded were those having a negative charge and high hydrophobicity such as diclofenac and ibuprofen.