(279e) Predicting the Extraction Behavior of Pharmaceuticals

Laube, F., Technische Universität Dortmund
Sadowski, G., Technische Universität Dortmund

Liquid-liquid extraction is a potential separation method for the purification and isolation of pharmaceuticals. Driving force of the liquid-liquid extraction is the different solubility of a pharmaceutical in two immiscible solvents (extraction system). Choosing the adequate extraction system and optimizing this system however usually requires a high experimental effort.

Therefore, within this work the PC-SAFT Equation of State [1] is used to predict the extraction behavior based on a limited number of solid-solubility data (SLE data) of the pharmaceutical in pure solvents. These data were used to determine the PC-SAFT pure-component parameters of the pharmaceutical as well as binary interaction parameters for the pharmaceutical/solvents systems. Solvent 1/solvent 2 binary parameters were fitted to binary liquid-liquid equilibria (LLE). Since binary LLE data for extraction systems can usually be found in literature, only the binary SLE data need to be measured. However, due to the fact that SLE data is also required for drug licensing and registration, this kind of data is in most cases available for pharmaceuticals.

Within this work the separation of different pharmaceuticals was investigated for varying extraction systems. Additionally the influence of pH and temperature on the extraction behavior was predicted. It was found that the modeling results of the extraction behavior are in good agreement with the experimental data. The sequence of solvent selectivity found from the modeling results was the same as the experimentally-determined one. Moreover the comparison of experimental data and modeling results shows that the influence of pH and temperature can be predicted correctly.

Hence, it can be concluded that the extraction behavior of pharmaceuticals can be successfully predicted by PC-SAFT based on their solubility in pure solvents, only. This drastically reduces the experimental effort for the selection of the solvent best suitable for extraction purposes.

1.            Gross, J. and Sadowski, G., Industrial & Engineering Chemistry Research 40 (2001) 1244-1260