WIENER FILTERING OF ELECTROANALYTICAL DATA BY MEANS OF FAST FOURIER-TRANSFORM

Stripping electroanalytical techniques, in spite of their high sensiti vity, usually suffer from poor signal-to-noise ratios. Both the numeri cal values of the ratio and the distribution of the noise across the r ecorded frequency spectrum must be considered carefully in order conve niently to separate the noise from the analytical signal. Traditionall y, this separation has been achieved by various filtering techniques t hat are time-consuming, require a visual and therefore subjective inte rpretation of the signal spectrum and have in general limited scope. I n this paper, a filtering technique that overcomes these limitations b ased on the Wiener filtering method in the frequency domain has been d eveloped. According to this scheme, the detection and separation of th e noise components in the voltammetric signal are automatically achiev ed in an optimal way. The principle of the method is that the noise in the signal can be estimated from the blank signal and subsequently el iminated. The feasibility of the proposed method for filtering in stri pping analysis was first assessed with simulated voltammograms and the n the technique was applied to real voltammetric data.