The development and optimisation of novel flow injection analysis methods for GSK
The purpose of this work was to exploit the technology of flow injection (FI) analysis to develop fast, new and automated novel methods for the benefit of the pharmaceutical and consumer healthcare (CH) industry at GSK. Two (general) approaches for optimisation of the method variables have been applied in this work. Classical univariate optimisation, where one variable is altered at a time (whilst keeping all others constant) was applied for the early publications in this thesis. A more recent and efficient approach however, is multivariate optimisation, in which two or more variables are altered simultaneously using the structured statistical (chemometric) technique called design of experiments (DOE). DOE was employed for the majority of method development in latter work, to perform a rapid screening to select the most critical variables and variable interactions; and then optimising these by modelling the effects on the responses in more detail. Multivariate investigation and optimisation was found to offer significant benefits for the efficient development of new methods. Several methods employing reaction with OPA (ortho-phthalaldehyde), developed and applied by the author, were further investigated/optimised in conjunction with the University of Brussels via a consortium project named ChemoAc (chemometrics in analytical chemistry). In these joint publications the original GSK methods were externally subjected to DOE optimisation and advanced chemometric techniques; in the latter papers with the objective of providing a generic methodology. This work has succeeded in developing several new and novel analytical methods, described in a total of sixteen publications and supporting analytical methodology enclosed within this thesis. Reviews of the technique of FI and the principles behind multivariate DOE serve to highlight the context of the research.