J.I. Drever S.R. Poulson L.L. Stillings 1998 Previous work on adsorption of oxalate at aluminosilicate surfaces suggests that maximum adsorption occurs through a bidentate attachment of the organic ligand, at near-neutral pH. Rates of ligand-promoted dissolution are expected to be greatest at this pH as well. We tested this model by measuring oxalate adsorption on the surface of andesine (An47), in solutions of pH 3- 5 and total oxalate concentrations of 0-8 mM. Contrary to expectation, the greatest adsorption density of 24 ??mol m-2 total oxalate was observed at pH 3 and 8 mM total oxalate. Adsorption is dependent upon the activities of both oxalate (C2O42-) and bioxalate (HC2O4-) in solution and can be modeled with either a two-term Langmuir or a two-term Freundlich isotherm. A Freundlich adsorption model provided the best fit to rate data because it was not constrained to a finite number of adsorption sites, as was the Langmuir model. The two-term ligand adsorption model was incorporated into a rate model: R(tot) = k(H-)[H(ads)/+](L) + k(HOx-)[HOx(ads)/-] + k(Ox2- )[Ox2(ads)/-] where R(tot) is the net dissolution rate of the feldspar, [i(ads)] is the concentration of species i adsorbed to the surface, and k(i) is the rate constant for release of the surface complex. The model was fit to data for oxalate-promoted dissolution of andesine, resulting in estimates for the rate constants of k(HOx-) = 1.16 x 10-12, k(Ox2-) = 1.05 x 10-12, and k(H-) = 9.61 x 10-13 mol of feldspar (??mol of i) (??mol of i)-1 s-1.Previous work on adsorption of oxalate at aluminosilicate surfaces suggests that maximum adsorption occurs through a bidentate attachment of the organic ligand, at near-neutral pH. Rates of ligand-promoted dissolution are expected to be greatest at this pH as well. We tested this model by measuring oxalate adsorption on the surface of andesine (An47), in solutions of pH 3-5 and total oxalate concentrations of 0-8 mM. Contrary to expectation, the greatest adsorption density of 24 ??mol m-2 total oxalate was observed at pH 3 and 8 mM total oxalate. Adsorption is dependent upon the activities of both oxalate (C2O42-) and bioxalate (HC2O4-) in solution and can be modeled with either a two-term Langmuir or a two-term Freundlich isotherm. A Freundlich adsorption model provided the best fit to rate data because it was not constrained to a finite number of adsorption sites, as was the Langmuir model. The two-term ligand adsorption model was incorporated into a rate model: Rtot = kH(+)[Hads+]L +kHOx(-) [HOxads-]+kOx(2-) [Oxads 2-] where Rtot is the net dissolution rate of the feldspar, [iads] is the concentration of species i adsorbed to the surface, and ki is the rate constant for release of the surface complex. The model was fit to data for oxalate-promoted dissolution of andesine, resulting in estimates for the rate constants of kHOx(-) = 1.16??10-12, kOx(2-) = 1.05??10-12, and kH(+) = 9.61??10-13 mol of feldspar (??mol of i)-1 s-1. application/pdf 10.1021/es980258d en ACS Oxalate adsorption at a plagioclase (An47) surface and models for ligand-promoted dissolution article