| Abstract: | The Influence of pH and Calcium Concentration on Metal-Fulvic Acid Bonds – Stripping Voltammetric and Chemometric Investigations. Electrochemically available metal concentrations of Cd, Cu and Zn were analyzed in dependence of different concentrations of calcium, fulvic acids and pH by Differential Pulse Anodic Stripping Voltammetry (DPASV). A statistical experimental design was the base to minimize the number of experiments and to include the multifarious interactions between the independent variables (Ca concentration, concentration of fulvic acids and the pH value) and otherwise the electrochemically available metal concentrations of Cd, Cu and Zn. At first analysis of experimental data was carried out by multiple linear regression. The main influence on available metal concentrations is the pH value especially for Cu. The Ca ion influence has not competitive effect and is only significant for Cd and Cu. It has an effect on the metal-fulvic acid-bonds. Two factor interactions exist for pH/fulvic acids and for Ca/fulvic acids too. The partial least squares regression (PLS) model was used to include the interactions between the metals Cd, Cu and Zn. For proving these two models the three parameters pH, Ca and fulvic acid were varied within the calibration range of the models and predicted values were compared with the experimental values. The approach with the PLS model is better than the approach with the multiple linear regression (normally used mathematical method of analysis of a factorial plan) with a relative error of 9.7% for modelling of Cd, 6.0% for Zn and 58% for Cu in relation to multiple linear regression with errors of 12% for Cd, 8.6% for Zn and 65% for Cu. The PLS modelling is a suited tool for modelling chemical interactions also in simulated natural matrices. |
