Article pubs.acs.org/est
Kinetic Signatures of Metals in the Presence of Suwannee River Fulvic Acid Jacqueline L. Levy,†,* Hao Zhang,† William Davison,† Josep Galceran,‡ and Jaume Puy‡ †
Lancaster Environment Centre, Lancaster University, Bailrigg LA1 4YW, United Kingdom Department de Química, Universitat de Lleida, Rovira Roure 191, 25198, Lleida, Spain
‡
S Supporting Information *
ABSTRACT: This work provides new information on the dissociation kinetics of metal-fulvic acid (FA) complexes. Diffusive gradients in thin-film (DGT) devices deployed in solutions containing metals and 30 mg L−1 Suwannee River FA at pH 5 and 7, at two different metal-to-ligand ratios, were used to estimate an apparent diffusive boundary layer (ADBL) thickness at the gel−solution interface. The discrepancy between the ADBL thickness measured for metals that are known to dissociate from complexes quickly (e.g., Cd) and that of other trace metals was exploited to calculate the rate of complex dissociation. When the ADBL thickness is plotted for a suite of metals, a “kinetic signature” is created. There was a clear kinetic signature at pH 7, with substantial kinetic limitation for Cu, Pb, and Ni and none for Cd, Co, and Mn (i.e., Cu-, Pb-, and Ni-FA complexes dissociated more slowly). At pH 5, the kinetic signature was less distinct, due in part to slow association kinetics of Mn, and possibly Cd and Co, with the resin. The good sensitivity of the method to small changes in dissociation kinetics was able to show that the dissociation of most metal-FA complexes is sufficiently fast to not limit the DGT measurement.
■
INTRODUCTION The chemistry of aquatic systems is rarely at equilibrium, but rather is dynamic, changing constantly with new inputs and with variations in physical parameters that can govern the interaction of contaminants. The interaction between trace metals and dissolved organic carbon (DOC) in the environment is critical for governing metal speciation and bioavailability.1,2 Metals bind to organic matter, forming metal complexes that vary in lability and availability.3,4 A system is termed (fully) labile when complex dissociation kinetics are so fast that a metal flux is limited by the diffusion of the complex. To assess the effect of lability, it is appropriate to use analytical techniques that are sensitive to it, such as diffusive gradients in thin films (DGT).3,4 DGT has been extensively used, in situ and in laboratory work, for measuring trace metals in freshwaters.5−7 A DGT device consists of a binding phase and a diffusion layer comprising a gel covered by a protective filter membrane, all in a protective plastic housing. Free metal ions and metal complexes diffuse through the polyacrylamide diffusive gel and filter. Metal is consumed by the resin binding layer consisting of a gel containing Chelex-100 resin.8 The strong binding of the metal to the resin ensures negligible net dissociation from the resin, allowing a concentration gradient to form through the diffusion layer. Metal−ligand (ML) complexes that are able to dissociate in the gel and filter domain are termed labile. The steady flux of metal from an aquatic system can be measured as the mass accumulated on the resin in a given time and then the labile fraction of metal in the solution can be calculated.8 © 2012 American Chemical Society
Varying the thickness of the diffusive layer of gel plus filter (g) changes the kinetic window in which in situ metal speciation measurements are made, i.e., when the diffusive gel is thicker there is more time for a given ML complex to dissociate and so the lability of the complex is effectively increased.9−11 This is particularly important when metal complexes with ligands like fulvic acid are present, as they are considered to be only partially labile using the standard DGT setup.10 In recent years, this concept of manipulation of the diffusive path length has been developed (including a full mathematical treatment) to provide information on the kinetics of dissociation of metal− ligand complexes.5,11,12 The use of multiple DGT devices with a range of diffusion layer thicknesses provides a measurement of the diffusive boundary layer (DBL) at the gel−solution interface of the DGT device.9,11,13 The thickness of the physical DBL, δ, decreases as the flow of water increases. At any given flow rate, it should be the same for all metals, as it is a common physical parameter. However, when the DBL is measured using DGT, it appears to vary for the different metals (so we refer to an apparent DBL or ADBL).11 Where dissociation of complexes is slow, the measured ADBL is larger than the physical DBL. Plotting the measured ADBL for a suite of metals provides a “kinetic signature” for that solutiona simple means of Received: Revised: Accepted: Published: 3335
December 1, 2011 February 17, 2012 February 18, 2012 February 21, 2012 dx.doi.org/10.1021/es2043068 | Environ. Sci. Technol. 2012, 46, 3335−3342
Environmental Science & Technology
Article
unreacted initiators. For thicker gels, longer washing times were required (up to 3 days).18 The final pH of the wash solution was measured, with values 2m is satisfied) the direct iterative procedure for 3339
dx.doi.org/10.1021/es2043068 | Environ. Sci. Technol. 2012, 46, 3335−3342
Environmental Science & Technology
Article
p The kinetic signatures and calculated gkin values at pH 7 show that dissociation of Cu, Pb, and Ni from FA are subject to the largest kinetic limitations for these experimental conditions. For p Cu gkin = 0.43 and 0.33 mm for the low and high metal-toligand ratios, respectively, while for Pb, they are 0.11 and 0.16 and for Ni 0.12 and 0.11, respectively. These findings agree with previous in situ measurements using DGT where Cu was found to be the most kinetically limited of the divalent trace metals tested, with only trivalent Al and Fe more subject to kinetic limitation.11,12 Of the metals tested in this study, Cu and Pb bind most strongly to FA and compete effectively with other metals for the stronger binding sites.2,12 This is reflected in the consistently high values calculated for the fraction of metal bound to fulvic acid (Table 1). As the binding strength increases, the rate of dissociation decreases according to the Eigen mechanism. Nickel does not compete as effectively as Cu and Pb for binding sites on FA, but Ni-FA has an intrinsically slow rate of dissociation.2,24 For Mn, Co, and Cd at pH 7 the ADBL corresponds with the theoretically predicted value of Pδ p and, consequently, there is no discernible gkin , consistent with no kinetic limitation. The ADBL values for Mn, in both the absence and presence p of fulvic acid, and the gkin values for Mn in FA solutions are particularly high at pH 5. This cannot be explained in terms of dissociation kinetics because, in the simple solution, only inorganic ligands are present (e.g., formation of free hydro and nitrate complexes), and all their complexes are likely to be fully labile, so that they fully contribute to the metal flux to the resin.8,25 Moreover, Mn is expected to have reasonably fast dissociation kinetics according to the Eigen mechanism.24 However, recent experimental evidence suggests that Mn binding to Chelex resins may be sufficiently slow to kinetically limit DGT measurements of Mn, particularly at pH ≤ 5.26 In this work, a higher ADBL of 0.37 mm was measured for Mn in FA-free media at pH 5 than for other metals (0.21−0.25 mm). This supports the idea that the kinetic limitation of Mn measurements when the pH is low (as assessed by high ADBL values) is actually due to slow binding to the Chelex. At pH 7, in the presence of fulvic acid, the ADBL values for Mn were low and very similar to values for Co and Cd, with values of 0.14 ± 0.03 and 0.16 ± 0.02 mm for the DGT devices deployed in the 3 μg L−1 and 9 μg L−1 trace metal solutions, respectively. The p values of 0.01 and 0.03 mm indicate little very low derived gkin or no kinetic limitation for Mn in the resin layer or in solution at this pH.8,25,27 p At pH 5, the kinetic effect is less distinct. The gkin values for Pb are similar to those at pH 7, but for Cu the kinetic limitation is much less marked than at pH 7. Within experimental error, there is little discernible kinetic limitation for Ni. Surprisingly, given their fairly weak binding, there is considerable kinetic limitation for Cd at pH 5 and for Co at pH 5 for the low metal case. Although a previous study did not detect slow binding of these metals to the resin at pH 5, the binding at pH 4 was slower than all other metals except Mn.26 Therefore, the possibility of slow binding to the resin, as for Mn, cannot be entirely ruled out. Metals preferentially occupy the strongest binding sites on the heterogeneous ligand, fulvic acid. As the concentration of metals in solution increases (and the metal-to-ligand ratio increases), increasingly weaker sites on FA are used to bind the metals. The metal bound to these sites will be more labile than the metal bound to the stronger sites on FA.12 Increasing the total metal concentration has the further effect of decreasing
the percentage of metal bound to fulvic acid and increasing the concentration of free metal (as shown with WHAM calculations of fraction M-FA for the current experiments; Table 1). Thus, as metal-to-ligand ratios increase, the potential for kinetic limitation would be expected to decrease. This p reasoning could explain the high gkin value for Co at pH 5 and low metal. It also fits well with the Cu data at both pH values p values at the higher metal-to-ligand where there were lower gkin ratio. However, for Ni and Pb at pH 7 and Cd at pH 5, this p values with increasing M:L ratio was decrease in ADBL and gkin not observed (Figure 2; Table 1). For all cases, it is likely that p the differences in gkin are within experimental error. The lability degree (ξ) expresses the ratio of the increase in the flux (with respect to that due to the free metal) associated with complex dissociation to the flux if the complex was fully labile.28,29 The complex penetration parameter, λML, expresses the effective penetration distance into the resin layer that would be necessary to ensure full dissociation of the complex.16 These two terms have also been calculated and displayed in Table 1. They show the greatest penetration for Co-FA complexes, which, for the lowest M:L ratio, are also least labile (ξ = 0.55). Generally, however, λML is Ni > Co > Mn > Cd ≈ Pb. When compared to this work for pH 7, the order for Cu, Ni, and Cd is the same, but the in situ value for Pb appears low, while Co and perhaps Mn values are high. These differences may be attributable to several factors, including different ratios of M:L, the presence in the water of ligands other than fulvic acid and the natural humic substances having different characteristics to that of extracted Suwannee River FA. The high ADBL values for Al and Fe in the field study dominate the signature. A study of kinetic signatures from metals complexed by algal exudates showed that Cu was again the most kinetically limited divalent metal, and again the ADBL value for Fe was much higher than any signal for divalent metals.14 A field study in a productive lake (unpublished data) suggests that Co complexes with algal exudates can dissociate slowly. Further work is underway to comprehensively assess the robustness of this “kinetic signature” approach, based on a broader, more diverse range of water quality parameters and to assess its potential diagnostic value.
■
ASSOCIATED CONTENT
S Supporting Information *
p The theory for derivation of the kinetic distance parameter, gkin , taking into account penetration of metal−ligand complexes into the resin and data on the kinetic signature of metals in simple solutions, i.e., when no fulvic acid is present (Figure SI-1) and p an example calculation of gkin (Table SI-1). This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: +1 418 654 3761; fax +1 418 654 2600; e-mail: jacqui.
[email protected]. Current address: Centre−Eau Terre Environnement, Institut National de la Recherche Scientifique, 490 rue de la Couronne, Québec, QC, Canada, G1K 9A9. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was funded by a Natural Environment Research Council Grant NE/E015859/1 (U.K.). Financial support from the Spanish Ministry of Education and Science (Projects CTQ2009-07831 and CTM2009-14612) and from the “Comissionat per a Universitats i Recerca del Departament d’Innovació, Universitats i Empresa de la Generalitat de Catalunya” (2009SGR00465) is acknowledged. We thank three anonymous reviewers for their helpful comments on the manuscript.
■
REFERENCES
(1) Filella, M. Freshwaters: Which NOM matters? Environ. Chem. Lett. 2009, 7, 21−35. 3341
dx.doi.org/10.1021/es2043068 | Environ. Sci. Technol. 2012, 46, 3335−3342
Environmental Science & Technology
Article
(21) Zhang, H. In-situ specation of Ni and Zn in freshwaters: Comparison between DGT measurements and speciation models. Environ. Sci. Technol. 2004, 38, 1421−1427. (22) Scally, S.; Davison, W.; Zhang, H. Diffusion coefficients of metals and metal complexes in hydrogels used in diffusive gradients in thin films. Anal. Chim. Acta 2006, 558, 222−229. (23) Scally, S.; Zhang, H.; Davison, W. Measurements of lead complexation with organic ligands using DGT. Aust. J. Chem. 2004, 57, 925−930. (24) Morel, F. M. M.; Hering, J. G. Principles and Applications of Aquatic Chemistry; John Wiley & Sons, Inc.: New York, 1993. (25) Gimpel, J.; Zhang, H.; Hutchinson, W.; Davison, W. Effect of solution composition, flow and deployment time on the measurement of trace metals by the diffusive gradient in thin films technique. Anal. Chim. Acta 2001, 448, 93−103. (26) Levy, J. L.; Zhang, H.; Davison, W.; Puy, J.; Galceran, J. Assessment of trace metal binding kinetics in the resin phase of diffusive gradients in thin films. Anal. Chim. Acta 2012, 717, 143−150. (27) Mundus, S.; Tandy, S.; Cheng, H.; Lombi, E.; Husted, S.; Holm, P. E.; Zhang, H. Applicability of diffusive gradients in thin films for measuring Mn in soils and freshwater sediments. Anal. Chem. 2011, 83, 8984−8991. (28) Galceran, J.; Puy, J.; Salvador, J.; Cecilia, J.; Mas, F.; Garces, J. L. Lability and mobility effects on mixtures of ligands under steady-state conditions. Phys. Chem. Chem. Phys. 2003, 5, 5091−5100. (29) Galceran, J.; Puy, J.; Salvador, J.; Cecilia, J.; Van Leeuwen, H. P. Voltammetric lability of metal omplexes at spherical microelectrodes with various radii. J. Electroanal. Chem. 2001, 505, 85−94.
3342
dx.doi.org/10.1021/es2043068 | Environ. Sci. Technol. 2012, 46, 3335−3342
