This paper presents generally shaped elution gradients and its impact on performance for preparative protein chromatography. The most common separation method in biopharmaceutical applications is preparative chromatography with step or gradient elution. In these cases the condition of the mobile phase is changed in steps or linearly in order to achieve the separation. In modern, precise and automated chromatographic systems it is possible to generate other types of elution gradients. In this work multi-step gradients are used to approximate generally shaped elution profiles and their pronounced impact on performance is discussed.<br/>General gradient profiles are found using model-based dynamic optimization. The model-based methodology rely on a mathematical model that can predict the separation and uses nonlinear programming (NLP) to find the optimal elution profile. The method uses discretization in space and time of the model and it solves the optimal-control problem by a primal-dual interior point method. The power of the methodology is discussed together with its current limitations.<br/>General gradients have substantial impact on performance due to the introduction of extra degrees of freedom in the separation problem. This can be used to achieve better separation, increased pool concentration, increased robustness and increased productivity and yield under given purity constraint. Model-based optimization of general gradients on a model system is compared with laboratory setup that can generate multi-step gradients. The impact of general gradients on performance is well illustrated experimentally on a tertiary protein ion-exchange separation problem.

General gradients for preparative chromatography

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.