A new method to quantify the CO 2 sensitivity of micro-organisms in shaken bioreactors and scale up to stirred tank fermentors

  • Eine neue Methode zur Quantifizierung der CO2 Empfindlichkeit von Mikro-Organismen in geschüttelten Bioreaktoren und deren Scale-Up auf gerührte Fermentatoren

Amoabediny, Ghassem; Büchs, Jochen (Thesis advisor)

Aachen : Publikationsserver der RWTH Aachen University (2006)
Dissertation / PhD Thesis

Aachen, Techn. Hochsch., Diss., 2006

Abstract

Small scale shaken bioreactors (e.g. shake flasks) traditionally equipped with different types of sterile closures are very useful tools in biotechnology. The gas transfer coefficient of the sterile closures (kplug) plays an important role in aeration of shaken bioreactors. The value of kplug depends on the average diffusion coefficient of oxygen (DeO2) and different lengths or/and diameters of the neck of flask. Therefore, in this study, a series of pipes with different lengths or/and diameters filled with cotton for a special shake flask, so-called ventilation flask, were employed. The gas transfer through the sterile closure of the ventilation flasks was characterized. Constant values of CO2 and O2 diffusion coefficient were found in all of the ventilation flasks. Considering these values and the neck geometry, a variety of kplug in ventilation flasks were obtained. Since, decreasing kplug causes a reduction of O2 concentration and an accumulation of CO2 in the gas phase of the shaken bioreactors, a realistic understanding and estimation of gas transfer in shaken bioreactors is advantageous to avoid oxygen limitation or carbon dioxide inhibition of a microbial culture. In this study, an unsteady state gas transfer model for shake flasks was developed and experimentally investigated for a wide range of gas transfer coefficients (kplug). The introduced approach is based on the model of Henzler and Schedel [23], which describes the spatially-resolved gas partial pressures inside the sterile closure, affected by the local gas diffusion coefficients and convective Stefan flow. For further easy processing, the resulting total mass transfer coefficient (kplug) is described as a function of the mass flow through the sterile plug (OTRplug) by an empirical equation. This equation is introduced into a simulation model which calculates the gas partial pressures in the head space of the flask. Additionally, the gas transfer rates through the sterile closure and gas-liquid interface inside the flask are provided. Simulations indicate that neglecting the oxygen in the head space volume of the flask at initial conditions (simple steady state assumption) may lead to an underestimation of the oxygen transfer into the liquid phase. The extension of error depends on the conditions. A good agreement between the introduced unsteady state model and experimental results for the sulfite and biological system confirmed the validity and usefulness of the proposed unsteady state approach. Moreover, a novel and easy method for quantification of CO2-sensitivity of microorganisms in ventilation flask was investigated, using the properties of ventilation flasks. The differences between the values of accumulated CO2 and concentration of oxygen in a culture system in ventilation flasks confirmed the validity of method. The effect of aeration on the removal of CO2 from the fermentation broth has been documented. Additionally, based on the data of the oxygen transfer rate (OTR), obtained by a Respiratory Activity Monitoring System (RAMOS) under a variety of specific aeration rates, the purposed new method was developed as an online monitoring method for CO2 sensitivity of microorganisms in shaken bioreactors.A maximum accumulated CO2 concentration of 12% was derived in both above methods, provided that the cultivation system is carried out under optimal conditions (e.g. the same filling volume (15m1), appropriate media and buffer capacity to control the pH, the suitable OTR (0.05 mol/l/h), operating under non oxygen limitation and RQ¡Ö1). The proper operation condition could be predicted using the unsteady state model. Applying these mentioned method, a significant effect of accumulated CO2 on the biomass concentration, growth rate and lysine product in the fermentation of C. glutamicum DM 1730 was found. Furthermore, the experimental results on Arxula adeninivorans LS3 and Hansenula polymorpha (WT ATCC 34438 and RB11-FMD-GFP) indicated that the CO2 had no effect on these microorganisms. Pseudomonas fluorescens DSM 50090 on yeast extract + glucose and Corynebacterium glutamicum ATCC WT13032 on L-lactate were found to be especially sensitive to CO2, which agree with literature. Some of the important advantages of the new methods are simplicity, lower cost and time consumption, easy of handling and producing similar results as large scale fermentation. Besides, a new aeration strategy from the ventilation flasks to an aerated fermentation system (e.g. measuring flask and stirred tank fermentor) was developed, based on the same concentration of gas compounds (O2 and CO2) in the headspace of these vessels. By applying this method, the concentrations of CO2 and O2 in the gas phase obtained from measuring (aerated) flasks and stirred tank bioreactors were comparable to those obtained from ventilation flasks. Finally in this study a new scale up method from shake flasks to stirred tank bioreactors, concerning the aeration strategy, was investigated based on the effect of CO2 ventilation. Even for different sets of aerations, similar trends were found for the values of the biomass concentration, L-lysine formation, maximum OTR and specific growth rate for fermentation of C .glutamicum DM 1730 as a model organism, in the both scales. Thus, the possibility of scaling up from ventilation flasks to stirred tank bioreactors based on CO2 ventilation criterion was demonstrated.