A voltammetric method for the quantitative determination of di- and threehydroxybenzene isomers including catecohol, hydroquinone and pyrogallol in three mixtures has been developed by recording differential pulse voltammograms between 100 and 400 mV with a bare and glassy carbon electrode modified with multi-wall carbon nanotube (GCE & MWCNT-GCE) in Britton-Robinson buffer of pH = 5.5. The chemical and instrumental variables affecting the analytical performance of the methodology were studied. It was found by applying the differential pulse voltammetric that nitrobenzene and three nitro-phenol compounds had well formed oxidation voltammetric waves with the peak potentials at 420, 381, 329 on GCE and 240, 145, 165 mV on MWCNT-GCE, respectively. For each compounds a linear calibration was obtained in the concentration range of 0.33-4.13, 0.33-5.94, 0.22-4.94 μgml-1 on GCE and 0.22-12.84, 0.22-11.18, 0.22-14.71 μgml-1 on MWCNT-GCE for catecohol, hydroquinone and pyrogallol, respectively. However, the voltammograms of these compounds are deeply overlapped and show nonlinear character which does not allow their direct determination without previous separation. The proposed method applies partial least squares regression (PLS) and artificial neural networks multivariate calibration to the resolution of this mixture using a set of potentials previously selected by genetic algorithms. The linear calibration graphs used to construct the calibration matrix were selected in the ranges from 0.44 to 9.46 μgml-1 for catecohol, from 0.44 to 7.11 μgml-1 for hydroquinone and from 0.50 to 7.03 μgml-1 for pyrogallol. A cross-validation procedure was used to select the number of factors. The prediction ability of the proposed method was tested with the use of a data set constructed from synthetic solutions for simultaneous determination of the three compounds, and in general, satisfactory results were obtained.
