Sequential response surface methodology for multioptimization in analytical chemistry with three-variable Doehlert designs

Abstract

In this paper, the use of Doehlert designs for the optimization of three experimental variables is proposed. Sequential response surface methodology is based on a strategy to obtain a quadratic function relationship between the response and the significant variables using a three and five-level experimental design. The Doehlert design structure allows us to study this reponse surface by estimating the parameters of the quadratic model and thus, by building sequential designs, it is possible to obtain the region in which the optimum values of such variables are simultaneously obtained. The mobility of these designs across the experimental domain of the variables permits the use of data from a previous design for a new design, obtained by response variation for the other two variables. The principle of this methodology is described and an example of its application in a spectrofluorimetric method is given.