(689b) Optimal Experimental Design in Plant-Material Extraction Considering the Process Costs

Plant-material extraction deals with the extraction of valuable components from natural raw materials. The kinetics and the achievable equilibrium of the extraction process can be improved be the application of pre-treatment methods like milling, drying or microwave treatment. Plant material contains a large number of different components of which often only a few are of interest. The selectivity and the yield of a process are strongly influenced by the polarity and the composition of the used solvent. Polar substances are in general extracted better by polar solvents and non-polar substances by non-polar solvents.

Due to the complexity of the plant material and the numerous influences on kinetics, equilibrium, a new approach to design plant-material extraction processes was developed. It is based on laboratory-scale experiments under different conditions. The process costs are determined considering fix (e.g. apparatus costs) and variable cost (e.g. price for plant material). Different experimental conditions will lead to different specific product costs. The optimal process is found, when the costs are minimized.

A systematic investigation of the pre-treatment methods and solvents requires a high number of laboratory experiments. Experiments are necessary for the modeling but costly and time-consuming. Therefore, an optimal experimental design method was used in this work. The Model-Based-Experimental Analysis (MEXA) is used to find the optimal conditions for the lab experiments. The tool MEXA fits extraction models to experimental data and determines the process conditions at which the specific process costs are minimal. Then those experimental conditions are found at which the error of the minimal product cost is smallest. This optimal condition will change, depending on the pre-treatment methods applied to the plant material. Thus the overall experimental effort is reduced while the specific product cost can be determined with a pre-defined accuracy.