The influence of data treatment on the interpretation of experimental results in voltammetry

In order to realize the advantages provided by recent improvements in data storage and treatment, enabled by computerized instrumentation, a procedure for data treatment in voltammetry is proposed. It is shown how the steps of this procedure are related and may affect the final conclusions. Examples of noise elimination, baseline subtraction and determination of peak potential are given. Concentration of the reactant in the simulated sample determined by the standard addition method can be hig hly influenced by the choice of baseline type. Depending on the peak determ ination procedure used and after the statistical tests provided, in a 4 mol l(-1) ionic strength media, models with three (beta(1) = 15.8 +/- 0.9, bet a(2) = 54.8 +/- 4.7, beta(3) = 171.5 +/- 3.8) and four (beta(1) = 14.5 +/- 0.4, beta(2) = 74.3 +/- 2,5, beta(3) = 129.2 +/- 3.2, beta(4) 11.2 +/- 1.0) labile lead (II) chlorocomplexes were obtained. Appreciating the more prec ise results, four stability constants are proposed. It is the shift of the detection limit towards lower values that allows one to prove the existence of the higher order complexes present in the solution, and it is obtained by more sophisticated software algorithms. The other evidence of the existe nce of the fourth complex is the statistically proven existence of four com plexes in a 6 mol l(-1) ionic strength media (beta(1) = 36 +/- 2, beta(2) = 232 +/- 39, beta(3) = 1360 +/- 105, beta(4) = 436 +/- 51), which is in acco rdance with the influence of ionic strength on the stability constants. (C) 1999 Elsevier Science B.V. All rights reserved.