As freshwater resources are predicted to become more scarce in the future, desalination will become a more prevalent treatment method. Microbial desalination, defined as the use of microbial processes to remove ions from a saline solution, may have potential as a new desalination method to produce water for domestic, agricultural or industrial purposes. This research explored the theory, performance, application and feasibility of microbial desalination. It consists of three research lines. Firstly, the performance and optimal conditions for microbial desalination were examined in batch experiments. In the second research line, the use of microbial ion transport proteins in a biomimetic membrane was proposed and the microbial light powered transport protein SyHr was genetically designed and expressed for this purpose. Finally, two models weremade which evaluate the feasibility ofmicrobial desalination at full scale; one which compares the costs of microbial desalination in sequencing batch reactors to seawater reverse osmosis and one to calculate the recovery and required size of a biomimetic membrane as described in the second
research line. Based on the results of all research lines, it was concluded that microbial desalination is be a promising new technology for desalination, which should be further developed.

Purple non-sulfur bacteria (PNSB) consist of wide genera of phototrophic bacteria found in various aquatic system. A high versatility in their mode of growth ranging from photoheterotrophic to dark fermentation gives them various potential applications. Their nature of being phototrophs require them to obtain light as energy source for growth and for that they require photopigments essential for light to ATP conversion. In contrast to pure culture of PNSB, the effect of different illumination on mixed culture PNSB is partly revealed. The response of altering light intensity on mixed culture of PNSB and the change of photopigments were to be answered in this study. Mixed culture of PNSB were grown in a sequencing batch reactor (SBR) operation of 8 h cycle. Four different light intensity settings were used to identify the effect of light on their growth and nutrient removal rate. Light distribution model was constructed to simulate the growth and nutrient removal of pure culture Rhodopseudomonas palustris in order to compare with mixed culture PNSB results. To track the growth and nutrient removal trend, biomass and nutrients (acetate, ammonium, and phosphate) concentrations at different points during reaction phase were measured. Bacteriochlorphyll content as major photopigment was analyzed with two different methods; micro well plate reader and reverse phased HPLC. It was shown that both growth rate and nutrient removal rate were significantly diminished as light intensity decreased. The decrease of growth rate with decreasing light intensity indicated that with the given conditions, light intensity was the growth rate-limiting
factor for the mixed culture PNSB studied. In comparison to simulation results, it was shown that the mixed culture PNSB was showing faster growth and higher removal rate. Bacteriochlorophyll content in the micro plate reader showed an increase to 75% intensity condition and then a decrease to 25% intensity conditions. Bacteriochlrophyll content analyzed by the two different methods showed relatively high deviation between each other as large error in both analyzing methods were present. The effect of changing light intensity on mixed culture PNSB with respect to its growth, nutrient removal and pigment content was studied. Further investigation on both higher and lower light intensity to identify the photo-inhibition level and saturation level is necessary. For a better approximation of the simulation to experimental data,
further works on precise measurement of light distribution in the photobioreactor is required. Photo-pigment analysis methods should be either replaced by a spectrophotometer using standardized cuvettes to minimize measurement error or be optimized to obtain reliable data.