Assessment methodologies for copper and zinc mobility in a neutral synthetic soil: The influence of pH

Speciation and redistribution of metals in soil are difficult to characterise experimentally. Nevertheless, environmental-impact assessment procedures are usually based on the tests developed to predict the mobility of the pollutant in different natural or artificial scenarios. In this work, a comparison between different experimental methodologies, including batch and continuous reactors and extraction tests, is carried out. Both batch reactors and laboratory columns are used, but on two slightly different scales. Sometimes, typical protocols have been modified in order to better explore accepted rules. For example, the current sequential extraction protocol was also applied in continuous mode. The study was carried out in a synthetic soil. The soil was composed so that it represented European neutral or acid using two relatively well-known metals, Cu2+ and Zn2+, soils (more problematic from an environmental point of view) but also taking into account the composition usually defined for eco-toxicological tests. The total retention of Cu2+ and Zn2+ obtained in column tests is normally lower than that obtained in batch systems. This result may be caused by precipitation. Cu2+ retention in continuous systems agrees well for both mono- and bimetallic systems. However Zn2+, which is not preferred by the medium, is displaced from the column by Cu2+, and its retention from the bimetallic solution is lower when compared with the monometallic system. Even when the measured outlet concentration of metal is equal to the feed solution, the solid/liquid system inside the columns is not usually in equilibrium: an axial profile of pH and metal concentrations is systematically obtained. The slow kinetic of carbonation reactions probably explains this behaviour, as determined by micro-column tests. Finally, standard extraction procedures applied in batch reactors do not lead to the same results when applied in continuous systems. Some reasons for this result are presented here, although their effects cannot be clearly quantified. (C) 2007 Elsevier B.V. All rights reserved.