Competitive ligand exchange voltammetric determination of iron organic complexation in seawater in two-ligand case: Examination of accuracy using computer simulation and elimination of artifacts using iterative non-linear multiple regression

Accuracy of the Scatchard linearization data processing method for competitive ligand exchange (CLE)–cathodic stripping voltammetry (CSV) measurements of seawater Fe-organic complexation in two-ligand case is examined with idealized Fe titrations data sets that are simulated using preset values of ligand parameters (conditional binding constants and total ligand concentrations). The results reveal substantial inherent artifacts for this method. An examination of patterns by which these artifacts vary with changes in Fe-binding strength of natural ligands relative to that of the added ligand suggests that the artifacts result not only from underestimated voltammetric sensitivity, but also from inadequate separation of individual ligand's contribution to Fe complexation at each titration point. For idealized simulated titration data, these artifacts can be eliminated by a procedure that combines non-linear regression with Turoczy and Sherwood's iteration. The method is demonstrated by reproducing preset values of ligand parameters over a diverse range of organic ligand alpha coefficients and by modeling titration data determined by CLE–CSV for seawater samples collected from the Eastern Bering Sea. Error analysis suggests that the sensitivity of model-derived ligand parameters to the error in the titration data is a strongly non-linear function of ligand parameters. Accurate measurement of titration data is thus required to use this method for accurate CSV measurements of seawater Fe-organic complexation.