In this 3 minute video, Didier Monnaie, PhD of Lonza Belgium introduces important considerations to improve a statistical design of experiment (DoE) for process chemistry.
The statistical design of experiment (DoE) method is a multivariate approach and aims to determine the relationship among factors affecting a process. It involves varying a number of potentially influential factors simultaneously. The design of experiment (DoE) method provides a better understanding of the cause and effect of process variability and leads to shorter development cycles, and it can also serve as the basis of the Quality-by-Design (QbD) approach which is of increasing importance.
In a design of experiment (DoE), the potentially critical factors need to be identified first, and controlled as tightly as possible during the experiments. If critical parameters are not controlled with sufficient accuracy noisy responses may result and the effects of the factors may not be visible. Subsequently, a less accurate design of experiment (DoE) study will require repetition of experiments. Depending on the reaction type and the nature of reactants,there are numerous critical factors that may be identified, such as temperature, stirring speed, addition rate, but also the concentration, catalyst type or amount, pH, pressure, etc.
Using an automated synthesis reactor, center points conditions can be repeated in identical experiments to validate the reproducibility of the experiment setup, control over process parameter, and provide confirmation of consistent analytical measurements. By proving that the reaction is repeatable, at these conditions scientists gain confidence in the precise control of the synthesis reactor and in situ analytical measurements. Once process confidence is established, no additional or repeat experiments are necessary to confirm the central composite design of experiment (DoE) set of process parameters.
Inaccurate temperature or dosing control will produce inconsistent results requiring repetition of the experiment in order to find the correct response. Conventional synthesis tools, such as round bottom flasks, do not meet the rigorous requirements of design of experiment (DoE) studies as the control of temperature varies with reactant addition. Often they also cannot cope with the heat release (or consumption) during the chemical reaction or the reactant addition resulting in undefined conditions.