Application of the hybrid genetic-simplex algorithm for deconvolution of electrochemical responses in SDLSV method

Abstract

A novel approach to the problem of deconvolution of highly overlapped voltammetric curves was proposed in this paper, which is based on a hybrid genetic algorithm. The hybrid genetic algorithm used in investigations was obtained by post-hybridisation and consisted of the simple genetic algorithm and the simplex downhill minimisation method. The algorithm designed in such a way inherited advantages of both search methods such as global search ability and accuracy typical for genetic algorithms and precision of the simplex method. It was successfully applied in simultaneous searching for appropriate values of six parameters of the model describing overlapping semidifferential voltammetric (SDLSV) curves. In the first step the algorithm was validated using a set of synthetic SDLSV curves generated for the Cd(II)/In(III) system and the results obtained were compared with those achieved by the simple genetic algorithm. After validation the proposed algorithm was used for deconvolution of SDLSV voltammetric curves obtained for the real Pb(II)/Tl(I) binary system for various ratios of ion concentration. The results presented prove that the proposed method is a very efficient deconvolution tool useful even in the case of serious overlap of electrochemical responses.