An experimental design approach to the optimisation of a flow injection analysis method for glycine

A flow injection analysis method for the determination of glycine, based on the reaction with ortho-phtalaldehyde and N-acetylcysteine in a basic buffer, was optimised. In the first step screening of the variables, to select the most important ones, was performed using: (i) a half-fraction factorial design and (ii) a quarter-fraction factorial design, for five factors at two levels. The effects of the factors on the peak height were calculated from both screening designs and compared. For the half-fraction factorial design (resolution IV), the significance of the factor effects on the peak height was checked by: (i) comparing them with a critical effect, calculated from two-factor interactions and based on a t-test, (ii) using a non parametric approach and (iii) drawing a normal probability plot. For the quarter-fraction factorial design (resolution III) the significance of the effects of the factors on the peak height was checked using: (i) a randomization test method, (ii) the non parametric method and (iii) a normal probability plot. In the second step, the factor found to be of importance was optimised using the uniplex method. (C) 1999 Elsevier Science B.V. All rights reserved.