Experimental design methodology and data analysis technique applied to optimise an organic synthesis

The study reported here was aimed at maximising the yield of a Michaelis-Becker dibromoalkane monophosphorylation reaction. In order to spare time and save money, we first applied a full factorial experimental design to search for the optimum conditions while performing a low number of experiments. In a second time, we used the Principal Component Analysis (PCA) technique to evidence two uncontrolled factors. Lastly, a special experimental design that took into account all the influential factors allowed us to determine the maximum-yield experimental conditions. This study also evidenced the complementary nature of experimental design methodology and data analysis techniques. (C) 2002 Elsevier Science B.V. All rights reserved.