M. Vega et al., APPLICATION OF THE TAGUCHI EXPERIMENTAL-DESIGN TO THE OPTIMIZATION OFA PHOTOOXIDATION PROCEDURE FOR TRACE-METAL ANALYSIS IN FRESH-WATER, Fresenius' journal of analytical chemistry, 350(3), 1994, pp. 139-144
A UV photo-oxidation device was constructed for the destruction of dis
solved organic matter (DOM) in freshwater samples in order to determin
e traces of zinc, cadmium, lead and copper bound to DOM as inert speci
es, by anodic stripping voltammetry in the differential pulse mode (DP
ASV). The photodigestor performance was optimised by means of the Tagu
chi experimental design. Four control factors (or design parameters) a
t three levels were explored: exposure-time to the UV-irradiation, pH
of the sample, hydrogen peroxide concentration and mineral acid added,
and assigned to the columns of a L(9)(3(4)) Saturated orthogonal arra
y. A noise factor (nitrilotriacetic acid concentration) at three level
s was introduced in the experiment to simulate the uncontrollable amou
nt of DOM in water samples. The experiment was conducted with two repl
ications of each trial and an optimum response insensitive to the vari
ations of DOM concentration was found when samples, acidified at pH 2
with sulphuric acid, were irradiated for 30 min in the presence of 12.
3 mmol/L hydrogen peroxide. The proposed procedure is more precise, ac
curate and fast than the wet digestion method.