(312g) A New Methodology for Estimation of Second Order Reversible Interaction Apparent Adsorption and Desorption Rate Constants of Protein Binding Kinetics Onto Resins in Packed-Bed Columns

In this study, a new and effective technique for characterizing apparent adsorption and desorption rates of protein binding kinetics in packed-bed columns is presented. Here second order reversible interaction between the protein and the resin is taken into account, where forward and reverse rate constants are k1 and k2 respectively. During the early stage of resin particle protein loading, desorption rate is usually negligible as compared with the forward adsorption rate due to the highly unsaturated nature of the resin. Thus, at the beginning of the column loading, second order reversible interaction adsorption desorption kinetics can be approximated by first order kinetics. This makes it possible to estimate k1 term from packed bed inlet and outlet stream concentration differences by the employment of standard chemical reaction engineering methodologies. Once forward rate constant k1 is known, by making use of the data obtained at prolonged times, then the calculation of reverse rate constant k2 becomes a straightforward task. Nevertheless, when the adsorption rate is low and/or the packed bed residence time is small, it is likely to come across with very small protein concentration differences between the bed inlet and outlet streams. Thus the measurement of the bulk liquid concentration differences between the column inlet and outlet locations, so as to find k1 and/or k2 can be very difficult and inaccurate. Due to the above mentioned limitations of once-through flow systems, for the model development, a packed-bed column with a recycle system is envisaged. This system allows the determination of packed bed protein adsorption desorption kinetics parameters, by mathematical interpretation of the experimental data gathered from a bench scale recycling flow packed bed adsorption column where cumulated time change of reservoir concentration is monitored. The experimentation can be continued as long as noticeable changes are observed in the reservoir protein solution concentration. During the experimental phase the apparent kinetic parameters of bovine serum albumin (BSA) onto a weak base ion exchange resin (Duolite A 568) were investigated. Reservoir BSA concentrations were monitored for different recirculation rates, column length/I.D. ratios and reservoir initial BSA concentrations. The governing mathematical equations of the recirculation system, based on kinetic rate constant model, were numerically solved by employing implicit finite differences method. The presented model is easily applicable to estimate second order reversible interaction apparent adsorption and desorption rates of protein binding kinetics in packed-bed columns for a wide range of protein solution concentration.