Annual Report 2020
Project Title: Intelligent decision system for efficient separations of molecular mixtures
The separation and purification of high-value chemical products (e.g., pharmaceuticals) from complex, multicomponent mixtures are crucial steps in their manufacture. To develop efficient manufacturing processes, it is essential that these steps minimise energy and material consumption, as well as waste production. Ultimately, accurate and reliable thermodynamic data, such as solubility, are required to achieve this goal.
My research is focussed on developing a novel methodology for determining accurate thermodynamic models of newly-discovered pharmaceutical compounds to aid design and development of separation and purification processes. The methodology is based on extensible thermodynamic models, which can be successively improved by addition of higher-order phenomenological models and incorporation of data from different experimental measurements. However, different measurements provide varying levels of information value at a cost related to the energy, effort, time and material required to perform the measurement. Therefore, the key challenge of my project is to present a method of identifying optimal sequences of experiments that provide enough data to correlate the thermodynamic models, yielding accurate descriptions of multicomponent pharmaceutical systems, which I hope to achieve using “machine learning”.
Awarded: Carnegie PhD Scholarship
University: University of Strathclyde