- 12:55 PM
79b

Practical and Rational Design of Bioseparation Processes Using Correlative Thermodynamic Models

Beckley K. Nfor1, Tangir Ahamed1, Marcel Ottens1, Emile J. A. X. Van der Sandt2, Michel Eppink3, Gijs W. K. van Dedem1, and Luuk A. M. van der Wielen4. (1) Delft University of Technology, Julianalaan 67, 2628 BC, Delft, Netherlands, (2) DSM Anti-infectives B.V., P.O.Box 425, 2600 AK, Delft, Netherlands, (3) DSP, Diosynth Biotechnology Europe, Oss, Netherlands, (4) Department of Biotehnology, Faculty of Applied Sciences and B-Basic, NWO-ACTS, Delft University of Technology, Julianalaan 67, 2628 BC, Delft, Netherlands

The number of bioprocess design alternatives grows dramatically when possible bioseparation methods are considered. Separation methods exploit different driving forces (gravitational, centrifugal, chemical potential, electric potential and pressure differences) that are rooted in the differences in molecular properties between the target compound and its contaminants. Industrial practice often involves intuitive qualitative concepts based on substantial experimental effort, creating many suboptimal situations. Hence, there is a need for thermodynamic data and/or predictive models, requiring an absolute minimum of experimental effort. We have developed several generalized models for predicting thermodynamic properties such as solubilities, activity coefficients, chemical potentials, and partition coefficients in solvent mixtures of small biomolecules from a few measurable thermodynamic parameters. These models allow translation of measurable properties, such as the second osmotic virial coefficient, B2, among different bioseparation techniques, facilitating rapid screening and design of unit operations. However, larger species such as proteins are more complicated because their phase behavior may be strongly affected by conformational changes. This paper extends the models to proteins and protein mixtures. Combined with high-throughput experimentation for rapid generation of thermodynamic data such as for B2, a framework and methodology will be created for systematic and faster bioprocess design.