Self-regenerating, suspended-phase whole-cell biosensor system employing micro-chemostat and cell engineering technologies

The university of Hertfordshire's role in this tri-group, dual university project was as follows:

i) develop the engineering for a novel small volume, no external moving part stirring system and associated control system for microbioreactor chambers;
ii) perform initial computational fluid dynamic modelling of the stirring process;
iii) develop compact new microbioreactor monitoring electronics to contribute to demonstrating the potential for portable systems in a military context.

The first objective has been met with the design and prototyping of a stirrer system employing a 1mm permanent magnetic ball that is free to move within, and sealed within, the microbioreactor chamber. The ball is actuated by four electromagnets whose energisation is sequenced so as to cause the ball to follow a trajectory through the fluid volume.
Whilst similar to a 'flea' stirring system, this is achieved without the presence of a rotating permanent magnet below the bioreactor. This is a key advantage in both simplicity, compactness and importantly, the ability to locate optical fibre reaction monitoring on the lower face of the bioreactor chamber.
The ball's movement results in both volumetric displacement and flow-eddies within the chamber resulting in a mixing process without the need for net fluid flow through the microbioreactor chamber. Stirring is required continuously as inhomogeneities of gas tension, metaboilites and nutrients dynamically develop within the essentially fixed volume as the result of the cell population's metabolic processes: such inhomogeneities will change local cellular growth conditions and may result in necrotic conditions.

Concerning modelling, initial work rapidly indicated that a full multiphysics model, encompassing magnetic interaction was beyond the resources of the project. Accordingly a validation simulation approach was adopted based upon CFD simulation of mixing in a chamber that is initially partitioned between two distinct fluids of identical properties. The ball was then stepped along the digitised trajectory of the ball in a high speed video of an experiment. This required an 'adaptive mesh' approach. The resulting model required a multi-day run time and significant computing resources, but showed a good qualitative agreement with the experimental video. This indicates that the adaptive meshing technique could be employed in the future as the basis of more advanced multi-physics simulations that could act as a design tool rather than a verification tool.

Finally, a single circuit board instrumentation system was developed replacing a number of distinct, interconnected bench-top instruments including 'lock-in' amplifiers. The compact unit provides for interrogation of the chamber's optical density (an index of cell population), Ph and dissolved oxygen. The unit was designed for low power operation, to the extent that it derived all its power from the USB bus. A cost reduction of circa one hundred times compared to the existing system was achieved. The unit was delivered to colleagues at UCL for evaluation.