Find the Set-Point Without Any Lab Work!
Process performance depends on strongly competing objectives, such as yield and purity, production rate, pool volume and pool concentration.
While a trial-and-error approach is unlikely to find a good compromise and Design-of-Experiments requires fractionation and offline analysis of all experiments, model-based optimization is almost for free! Within seconds, a chromatogram is generated and all fractionation possibilities are evaluated with respect to all chosen objectives.
The optimization procedure either works with a weighted sum of sub-objectives or as a multi-criteria optimization. The latter approach is shown in the adjoining plot. Here, the Pareto surface for yield, purity and production rate was identified using a genetic algorithm.
Simulations (black dots) generated and evaluated with ChromX.Pareto surface visualized with Origin 8 by OriginLab Corporation.