Computational Fluid Dynamics (CFD) has become a pivotal tool in the design and scale-up of bioprocesses by providing deep process insights. It has been extensively researched and used in stirred tank reactors. Research for shaken vessels, like shake flasks, especially at elevated viscosity, is, however, very limited. In this work, a CFD model for 250 mL shake flasks has been established, simulating a variety of shaking conditions, including viscosity of up to 100 mPa∙s. CFD model results were then used to calculate volumetric power input, volumetric gas-liquid mass transfer rate (kLa) and shear rates. Simulated fluid flow agrees well with experimentally recorded liquid distributions. Calculated volumetric power inputs deviate on average by 0.18 kW/m3 from experimentally determined volumetric power inputs. kLa values, spanning almost three orders of magnitude, deviated by less than a factor of two from published data for kLa values in shake flasks. Shear rates were calculated in two ways, where the first showed unsatisfactory results. The second approach, which is based on the largest eigenvalue of the strain rate tensor, was on average 16 % smaller than indicated by published data for effective shear rates in shake flasks. Lastly, the CFD model highlighted the inherently gradual nature of the transition between so-called in-phase and out-of-phase operating conditions in shake flasks. Process monitoring is also crucially important for design and scale-up of bioprocesses. Traditional offline sampling is, however, highly laborious, a risk for contamination and provides limited time resolution. Hence, online monitoring of key process parameters is highly advantageous. In this work, online monitoring of dissolved oxygen tension (DOT), pH, scattered light and viscosity were combined in a single monitoring system for shake flasks, termed ShakeVisc. DOT measurements were performed with oxygen sensitive fluorescent nanoparticles. The pH was measured with pH sensitive fluorescent sensor spots. Scattered light was measured at 610 – 630 nm with a newly developed optical sensor. The shift of the angle of the bulk liquid is correlated to the viscosity. This shift was determined based on the fluorescence intensity of the oxygen sensitive nanoparticles and, for the first time, based on scattered light signal intensity. The ShakeVisc system was successfully applied to cultivations of Escherichia coli, Paenibacillus polymyxa and Xanthomonas campestris.
Carl Dinter (Wed,) studied this question.