Organic complexation of Al and Fe in acidic soil solutions: Comparison of diffusive gradients in thin films analyses with Models V and VI predictions

Abstract

The last decade has seen important development of analytical and modeling tools to investigate the interactions in solution between metals and dissolved organic matter. We evaluated the internal consistency of the analytical method of diffusive gradients in thin films (DGT) and the speciation Models V and VI for investigating organic complexation of Al and Fe in acidic sandy soil solutions. For this purpose we used our dataset obtained by DGT of the speciation of Al and Fe(III) between dissolved inorganic metal and soluble metal-dissolved organic matter (DOM) complexes in acidic soil solutions at pH=3.5, 4.0 and 4.5 and different imposed metal/organic carbon (M/C) ratios in solution. We modeled this data-set with Models V and VI, using the default parameter values but optimizing the fraction of DOM that is inert with respect to metal binding (24%). The calculated speciation of Al by both Models V and VI was generally in good agreement with the DGT results. However, at low Al/C ratios the Al fraction measured by DGT was higher than the dissolved inorganic fraction calculated by the models, most likely due to the detection of small (<1000 Da), labile Al–DOM complexes by DGT. This discrepancy did not occur with Fe(III) probably because its complexes with DOM are not labile enough to be detected by DGT. The dissolved, inorganic Fe fractions calculated by Model V were in good agreement with the DGT results, especially considering the fact that extra uncertainty was introduced by the analytically determined Fe(II)/Fe(III) speciation in solution. Model VI calculated the trends in Fe binding well, but generally estimated a higher dissolved, inorganic Fe fraction than Model V and DGT. The Fe(III) binding constant in Model VI was recently modified to reflect stronger binding to DOM. Since both Models V and VI predicted the inorganic Fe in solution to consist mainly of Fe(II), we propose a similar critical evaluation of the Fe(II) binding constant in Model VI.