Field-flow fractionation characterization and binding properties of particulate and colloidal organic matter from the Rio Amazon and Rio Negro

Colloidal and particulate natural organic matter was fractionated and concentrated from the Amazon and Rio Negro rivers using tangential-flow filtration (TFF). Flow field-flow fractionation, with UV absorbance detection, was used to investigate the molecular weight distributions of the organic colloids. To further characterize the nature of the Rio Negro colloids, the size distributions of the iron and aluminum concentrations were determined by off-line graphite furnace atomic absorption spectrometry analysis of fractions collected during the FFF separation. The size distributions obtained by FFF were considerably smaller than expected from the stated pore size of the TFF membranes. These results demonstrate that care must be taken in using TFF to classify the size distribution of organic colloids and associated elements present in rivers. Total metal complexing capacities (TMCC) of the colloids were determined by voltametry at various pH using lead as a probe. Total binding capacities were measured by automated proton titration of the particulate and the colloidal fractions. To describe Pb binding together with Al competition and the presence of mineral forms of Al in the particulate fraction, the NICA-Donnan approach was used. The fitted model lines agree reasonably well with the experimental results, suggesting that natural ligands behave like purified humic or fulvic acids. This result is surprising due to the extreme treatment humic substances are subjected to during purification. The only major difference between the binding constant values obtained for purified humic substances and the TFF fractions is a higher affinity constant value for the carboxylic groups of the latter. An interesting result is the agreement in the metal binding for the different fractions at pH 4 and 5. which implies that the type of groups present in these fractions are similar. (C) 2002 Elsevier Science Ltd. All rights reserved.