Application of the split-plot experimental design for the optimization of a catalytic procedure for the determination of Cr(VI)

A kinetic catalytic procedure for the determination of Cr(VI), based on the o-dianisidine oxidation reaction with hydrogen peroxide in a weakly acid medium was optimized with respect to both reactants [HCl, o-dianisidine and H2O2] and solvent composition [mixtures of water, acetone and N, N-dimethylformamide], in order to achieve higher sensitivities. This was accomplished by making use of a split-plot experimental design which permitted simultaneous variations in both mixture and process variables, using a response surface approach. Ten mixture experiments were randomized at each one of the 23 factorial levels and error analysis was performed using a split-plot approach. The optimal mixture [water-acetone 70-30% m/m] and process variables [HCl: 6.0 x 10(-4) mol/l, o-dianisidine: 1.9 x 10(-2) mol/l and H2O2: 0.79 mol/l] values resulted in a significant sensitivity increase compared with a similar procedure described in the literature. (C) 1998 Elsevier Science B.V. All rights reserved.