Cwk. Chow et al., A NEURAL-NETWORK APPROACH TO ZINC AND COPPER INTERFERENCES IN POTENTIOMETRIC STRIPPING ANALYSIS, Journal of intelligent material systems and structures, 8(2), 1997, pp. 177-183
Zinc and copper are two elements which mutually interfere with each ot
her in stripping analysis. The cause is the formation of a Zn-Cu inter
metallic compound in the mercury film, which affects both Cu and Zn an
alyses. A backward error propagation artificial neural network has bee
n applied in a novel approach for the determination of zinc in the pre
sence of copper using potentiometric stripping analysis. This performe
d well in determining the correct zinc concentration in the sample whe
n provided with the stripping times of zinc and copper and the copper
concentration (determined by shifting the plating potential to a lower
value to prevent the zinc being plated onto the mercury film electrod
e). The unknown zinc concentration was determined following an initial
period of network exposure to a set of experimental data, which were
used as examples of the required input/output data mapping.