Abstract

The voltammetric behavior of cinnamic acid (CA) and 3, 4-dihydroxy benzoic acid (BA) at an glassy carbon disc electrode has been investigated. The obtained cyclic voltammograms showed that their oxidation peaks were overlapped and it was difficult to determine them individually from a mixture without separation. To overcome this limitation, partial least squares (PLS) modeling, as a powerful multivariate statistical tool, was proposed for the resolution of the overlapped voltammetric signals from mixtures of CA and BA. With the help of the orthogonal signal correction (OSC), the unrelated information was removed and the results were confirmed. In this procedure, the genetic algorithm (GA) was used for the potentials selection for PLS. In addition, OSC-GA-PLS for the first time was used and the results of this method were better than those of the other methods. The proposed method was suitable for the determination of CA and BA in mixtures.

Keywords: Chemometrics, Genetic algorithm, Orthogonal signal correction, Partial least squares, Cinnamic acid, 3,4-Dihydroxy benzoic acid