Production of succinic acid by Actinobacillus succinogenes in attached-growth bioreactors and dynamic modelling of biofilm formation

Sammanfattning: Actinobacillus succinogenes is a microorganism is a very efficient producer of succinic acid, a chemical with a wide range of industrial applications. The bacterium is prone to form biofilm, a flocculation of cells attached to a surface. In this study, the bioproduction of succinic acid using Actinobacillus succinigenes 130Z was performed in microplate, bottle, and bioreactor systems to gain new insight regarding the biofilm formation process. From microplate experiments the growth kinetics of both suspended and attached cells were characterized at different initial pH and concentration of yeast extract, indicating that biofilm formation is enhanced at lower pH. The maximum specific growth rate was determined to be between 0.178 h-1 and 0.413 h-1 for suspended cells and 0.063 h-1 and 0.089 h-1 for attached cells, depending on the initial conditions. During batch fermentation in glass bottles, Kaldnes K1 packing material was used as biofilm support material for comparison between fermentation without and with packing material. No significant difference was found regarding suspended cell growth, substrate consumption, and product formation. A difference was observed regarding the formation of biofilm as measurements indicated more rapid biofilm formation and elevated levels of biofilm on packing material compared to glass surfaces. The maximum specific growth rates when no packing material was present were 0.224 h-1 and 0.413 h-1 for 20 ml bottles and 100 ml bottles, respectively. With packing material these values were 0.209 h-1 and 0.427 h-1. Kinetic parameters calculated from the experiments were used to calibrate a kinetic model which well represented the experimental results. A bioreactor setup using a bubble column reactor was constructed and fermentations were performed in both batch and continuous operation. The bioreactor experiments were all contaminated at some time during the fermentation processes with limited results. Using data from a CSTR fermentation the model was validated which showed difficulties in predicting suspended and attached cell growth.