DEVELOPING A KINETIC ALTERNATIVE IN MODELING SOIL ALUMINIUM MATTIAS ALVETEG, HARALD SVERDRUP and PER WARFVINGE Department of Chemical Engineering II, Lund Institute of Technology, Lund, Sweden (Received August 5, 1994; accepted August 18, 1994) Abstract. Soil chemistry models often use gibbsite solubility and similar equilibrium models to predict A1 concentrations in soil solution. A kinetic alternative was developed with the goal of finding universal rate constants instead of the site- and depth-specific solubility constants usually associated with the equilibrium approach. The behavior of the two approaches was studied within the framework of the steady-state soil chemistry model PROFILE using data from Soiling, Germany and G/irdsjOn, Sweden, two sites with different mineralogy and land use history. The kinetic alternative uses a mass balance to predict A1 concentrations. The sources of A1 in soil water are deposition, weathering and mineralization. The sinks are leaching and the formation of an aluminosilicate precursor. The precursor slowly transforms into an ordinary clay mineral. Both formation and transformation of the precursor are treated as irreversible processes. The kinetic model introduces a new relationship between pH and A1 and produces a systematic pattern of different apparent gibbsite equilibrium constants at different depths. Results show that the kinetic model systematically underestimates A1 concentration in the upper horizons, which indicates that there may be additional sources of A1 in the upper horizons not accounted for in the model. Predicted values of pH and A1 concentrations are comparable with field observations. 1. Introduction Soil acidification and associated high A1 concentrations are today one of the most important environmental problems in northern Europe. Evidence has been present- ed that elevated A1 concentrations decrease forest growth (Sverdrup and Warfvinge, 1993) and it is therefore important to develop tools to predict A1 concentrations in acidified soils. The objective of this paper is to study the possibility of modeling soil A1 with a kinetic approach using a mass balance for A1. The kinetic approach was tested using data from two sites with quite different mineralogy, the Soiling site in Germany and the G~rdsj6n site in Sweden. The A1 concentration in soil water is often modeled assuming an equilibrium between A1 in solution and A1 in the solid phase. A common assumption is that the solid phase controlling the A1 concentration is gibbsite. The gibbsite solubil- ity constant, relates the concentration of free A13+ to pH through [A13+] = Kg [H+] 3. Gibbsite solubility, however, has very slow kinetics (Bloom, 1983), and it is therefore highly improbable that the A1 concentration in a solution in contact with AI(OH)3 will be governed by an equilibrium. This is also demonstrated by the wide range of apparent solubility constants found in field measurements (see Table I) which cast doubt on the possibility of predicting A1 concentrations with gibbsite solubility constants. Apart from these problems with the gibbsite solubility Water, Air and Soil Pollution 79: 377-389, t995. 9 1995 KluwerAcademic Publishers. Printed in the Netherlands.