Optimization of on-line microwave flow injection analysis system by artificial neural networks for the determination of ruthenium

A methodology based on the coupling of experimental design and artificial n eural networks (ANNs) was proposed in the optimization of a new on-line mic rowave flow injection system (FIA) for the determination of ruthenium, grou nded on its catalytic effect on the oxidation of dibromocarboxyarsenazo (DB M-AsA) by potassium periodide under the microwave irradiation. The response function (RF) used was a weighted linear combination of two variables rela ted to sensitivity and sampling rate. A neural network with extended delta- bar-delta (EDBD) learning algorithm was applied to predict the maximal RE a ccording to which the optimized conditions were obtained. The optimized new on-line microwave FIA system is able to determine ruthenium in 5-200 ng ml (-1) range with a detection limit of 2.1 ng ml(-1) and a recovery of 94.6%. A sampling rate of 58 h(-1) was obtained. In contrast to traditional metho ds, the use of this methodology has advantages in terms of a reduction in a nalysis time and an improvement in the ability of optimization. (C) 2001 El sevier Science B.V. All rights reserved.