A Model-Based Approach to Determine the Design Space of Preparative Chromatography

A dream of many pharmaceutical companies is to be able to register a large design space with the regulatory agencies. The problem is that this will require both time and money, so an intelligent method of validating a design space is needed. The design space should only cover operating points at which the process runs optimally. This means that the process should be optimized for different process scenarios and objective functions and the found operating points should be registered as design space. This paper presents a method of determining a good design space by creating Pareto fronts for the ideal case and for various process disturbance scenarios. Optimal operating points are found for varying ratios between feed costs and operating costs, making it possible to make a quantitative choice of an operating point based on this ratio and a qualitative choice based on the whole front. The analysis will show how the chromatographic process can be made more robust when optimizing for higher yields, and how the effect of the critical process parameters can change. To be certain that a robust process is found and that it has a high performance, process disturbances must be taken into account when optimizing a process