(370a) A Validated Mathematical Model for the Simulation of Affinity Membrane Chromatography Process

Boi, C., University of Bologna
Sarti, G. C., University of Bologna

Membrane affinity chromatography is seriously considered as a possible alternative to bead based chromatography. Membrane chromatography is not limited by diffusion as the majority of chromatographic resins and therefore it is particularly suited for purifying large bio-molecules such as high molecular mass proteins and viruses. In order to promote its applications in up-scaled plants, it is imperative to develop a reliable simulation tool able to describe the process performance in a predictive way.

In this work we present a mathematical model that can describe a complete affinity cycle decoupled in its chromatographic steps: adsorption, washing and elution. The mathematical formalization takes into account all the relevant mass transport and kinetics phenomena involved in the membrane affinity chromatography process, namely axial convection, longitudinal dispersion in the micro-porous matrix and affinity binding with the specific adsorption site. In addition, also extra-column effects resulting in mixing volumes and delay time are included in the model and their parameters are determined through independent experiments .

The few relevant fitting parameters were derived from a calibration with an extensive set of experimental affinity cycles performed with pure IgG solutions. The affinity cycles have been carried out using different innovative affinity membranes tested under a broad spectrum of operating conditions.

Model validation is then achieved by comparing simulation results with experimental data which have been obtained for the purification of immunoglobulin G from a complex feed as a cell culture supernatant.

Model simulations are in good agreement with the experimental affinity cycles both in the case of pure IgG solutions as well as for the cell culture supernatant considered which contains monoclonal hIgG. Remarkably, in the latter case the model is entirely predictive, since the same values of the parameters used for pure IgG solutions are sufficient to describe the affinity chromatography cycle, thus demonstrating the solidity of the model presented.


This work has been performed as part of the ?Advanced Interactive Materials by Design? (AIMs) project, supported by the Sixth Research Framework Programme of the European Union (NMP3-CT-2004-500160).