Staged Dataset for Design of Experiments (DoE)

P. Allan, A. Nordon and C. Jamieson, University of Strathclyde

Understanding and optimising reactions is vital in the chemical and pharmaceutical industry to improve product quality and reproducibility. In addition it helps to minimise development costs, batch time and waste. Design of experiments (DoE) is a process used in many industrial and pharmaceutical applications to gain a better understanding and in turn find the true optimum of a process. The use of DoE is slowly increasing in academia, however it is not fully utilised in the undergraduate curriculum or as part of research projects. One reason for this is the lack of suitable data sets available to use for an undergraduate experiment. With increasing student numbers there is a shortage of time and resources to carry out the quantity of practical experiments required for a DoE in an undergraduate teaching laboratory.

A computer based DoE experiment was successfully introduced into the 3rd year undergraduate curriculum at the University of Strathclyde. The students did not complete any practical experiments and instead were provided with a complete data set previously generated from an amide bond formation reaction. The students manipulated the data using the software Design Expert® to go through the processes of scoping, screening and optimisation. Overall positive feedback was obtained from a Likert style questionnaire completed by students. To date over 70 students have carried out the experiment with over 94 % of students reporting the experiment was interesting, valuable to their studies, increased their understanding of DOE, allowed them to make the link between analytical chemistry and the other branches of chemistry, and they will now consider the factors involved in optimisation when carrying out other experiments. One improvement suggested by students was to ‘optimize the experiment then carry it out.’ The conditions in which this particular reaction was carried out did not lend itself well to being performed by 3rd year undergraduate students in a short time frame. To address the students feedback and bridge the gap a staged data set from a relevant organic synthetic reaction has been produced that can then be utilised for a guided computer based DoE in an analytical undergraduate teaching laboratory and also allows the students to carry out the optimised reaction in the organic teaching laboratory.

The model Suzuki-Miyaura cross coupling reaction was used to generate an appropriately staged DoE dataset for an undergraduate cohort to interrogate further. The Suzuki-Miyaura reaction2 was chosen as it is relevant and familiar to undergraduate students with the theory being taught in their second year organic lectures at the University of Strathclyde. The optimised reaction conditions are also suitable for a practical 2nd year undergraduate laboratory experiment as it can be carried out in water (cheap, non-toxic and non-flammable) within a reasonable time frame and it can be considered a “green process”. To complement the analytical lectures on the theory of DoE introducing a guided computer based DoE in the Analytical teaching laboratory based on a familiar organic reaction importantly allowed students to make the link between organic and analytical chemistry. The creation of the staged data set allowed a computer based DoE study to be created and completed in the Analytical teaching laboratory with the option of a practical element that can be carried out by undergraduate or postgraduate students in the organic teaching laboratory, increasing their DoE knowledge.