Luuk K. Koopal

Ion binding to natural organic matter : competition, heterogeneity, stoichiometry and thermodynamic consistency

Abstract The general principles of cation binding to humic materials are discussed. Important aspects that need to be included in general purpose speciation models are: the extreme binding heterogeneity of natural humic materials, the variable stoichiometry of binding (monodentate, bidentate and tridentate), the competition between specifically-bound ions, especially protons and metal ions, and electrostatic effects which give rise to ionic strength effects and the non-specific binding of counterions. The NICCA–Donnan model is a semi-empirical model that addresses these issues. It is similar to the previously published NICA–Donnan model except that it introduces an additional degree of scaling that ensures thermodynamic consistency and allows for variable stoichiometry of binding. It implicitly accounts for the large degree of chemical heterogeneity of humic particles. The NICCA (consistent NICA) model also recognizes that the affinity distributions are ion specific and are not fully correlated. The model requires no assumptions to be made about the geometry of the humic particles, but the Donnan submodel does allow for shrinking and swelling. Important model parameters such as the site density and median binding constants ( log K ) are not dependent on pH, metal ion concentration, ionic strength, etc. Data are analysed for H+, Ca2+, Cd2+, Cu2+, Pb2+ and Al3+ binding to a single purified peat humic acid. The NICCA–Donnan model captures the non-linearity of the observed isotherms even at very low free metal ion concentrations. After fitting the model to datasets containing only the proton and one metal ion, the model was able to predict Cd2+–Ca2+, Cu2+–Ca2+ and Pb2+–Al3+ competition reasonably well. It also gave satisfactory predictions of the H+/Mz+ molar exchange ratios. These ratios varied strongly with metal ion: Ca2+ (0.2–0.5); Cd2+ (0.5–1.0); Pb2+ (1.1–1.2); Cu2+ (1.2–1.7) and Al3+ (2.1–2.7), and also to a varying degree with pH and free metal ion concentration.

Metal ion binding to humic substances: application of the non-ideal competitive adsorption model.

The application of a new model to describe metal ion binding by humic acids is discussed. Metal ion binding is always of a competitive nature since the proton is always present. Although of great practical importance, the combination of a chemically heterogeneous system with competitive binding poses difficult problems from both experimental and theoretical points of view. The new Non-Ideal Competitive Adsorption model (NICA model) used here is able to account for the non-ideal binding to heterogeneous ligands. A good description of the binding of H, Ca, Cd, and Cu to a purified peat humic acid is achieved over a wide range of free metal ion concentrations (-2 > log Me 2+ > -14) and pH (2 < pH < 10). The results show that binding of metal ions to humic acid is strongly influenced by the intrinsic chemical heterogeneity of the humic material itself as well as by ion-specific non-ideality. The results indicate that copper competes much more efficiently with protons bound to the phenolic type groups than calcium and cadmium.

Measurement and interpretation of electrokinetic phenomena - (IUPAC technical report)

In this report, the status quo and recent progress in electrokinetics are reviewed. Practical rules are recommended for performing electrokinetic measurements and interpreting their results in terms of well-defined quantities, the most familiar being the z-potential or electrokinetic potential. This potential is a property of charged interfaces, and it should be independent of the technique used for its determination. However, often the z-potential is not the only property electrokinetically characterizing the electrical state of the interfacial region; the excess conductivity of the stagnant layer is an additional parameter. The requirement to obtain the z-potential is that electrokinetic theories be correctly used and applied within their range of validity. Basic theories and their application ranges are discussed. A thorough description of the main electrokinetic methods is given; special attention is paid to their ranges of applicability as well as to the validity of the underlying theoretical models. Electrokinetic consistency tests are proposed in order to assess the validity of the z-potentials obtained. The recommendations given in the report apply mainly to smooth and homogeneous solid particles and plugs in aqueous systems; some attention is paid to nonaqueous media and less ideal surfaces.

Metal ion binding by natural organic matter: from the model to the field.

With the newly developed NICCA-Donnan model, we estimate the activity of toxic metal ions from simple measurements like total metal concentration and organic matter content. The model evaluates Cu and Cd binding from three field systems, a mountain lake and two sandy soils, using model parameters calibrated for natural organic matter analogues with laboratory measurements. This is possible because the model includes site binding heterogeneity, electrostatic effects, competitive binding, and ion specific nonideality. The predictions derived closely matched the field observations when site binding densities are adjusted. The partition coefficients between soil and soil solution were also predicted for Cd and Cu under conditions where the organic matter controls the metal binding in both soil and soil solution. The model calculations show that in soil solutions 50% of the Cd and 99.99% of the Cu is bound to the dissolved organic matter. The model can be used to evaluate the effects of variations in the chemical conditions (e.g., acidification or total metal loading) on the free metal ion concentration in solution.

Electrolyte adsorption on heterogeneous surfaces: adsorption models

Abstract Several heterogeneous electrolyte adsorption models are derived. The underlying assumptions for these models and their applicability are discussed. It is shown that the pristine point of zero charge (PPZC) may well be a function of the degree of surface heterogeneity. On the other hand model calculations show that the shape of the surface charge (σ0)-pH curves is very insensitive towards the degree of heterogeneity. It thus follows that σ0-pH curves for metal oxides, exhibiting some degree of heterogeneity, may be described with a homogeneous electrolyte adsorption model. Interestingly it appears that the titration curves of TiO2 in KNO3 can be described equally well using a “one-pK” model as with a “two-pK” model.

Metal ion adsorption on heterogeneous surfaces; Adsorption models.

The sensitivity of metal ion adsorption on metal oxides for the degree of random surface heterogeneity (smeared out potential) using three electrolyte adsorption models is investigated. It is shown that the effect of random surface heterogeneity is hard to detect when the metal ion adsorption constants are adjusted together with the adjustment of Kq, i.e., a shift in the position of the energy distribution curve. This shift is necessary in order to keep the experimentally determined pristine point of zero charge at a fixed position. Some reasons for the low sensitivity are indicated. Cadmium adsorption isotherms for hematite and amorphous iron oxide in combination with CdH exchange data obtained at several pH values can be described satisfactorily with a simple homogeneous surface one- or two-pK basic Stern model. For both iron oxides the predominant surface species of cadmium is SO−CdOH+.

Analysis of proton binding by a peat humic acid using a simple electrostatic model

Detailed potentiometric titration data were collected for a purified peat humic acid (PPHA) over a range of pH (pH 3.5–10.5) and KNO3 background electrolyte concentrations (0.001–0.3 M). The data were analyzed following the master curve approach which includes both an electrostatic double layer model and a model for the intrinsic heterogeneity of the PPHA. Spherical and cylindrical double layer models gave equally good fits to the data. A salt dependence observed around pH 5 could not be completely removed by taking into account the electrostatic interactions. Hysteresis was observed to a much greater extent in the first titration cycle compared with the second cycle. This suggested that some slow and only partly reversible aggregation was occurring possibly as a result of the aggregation created during the purification of the humic acid. Titration curves for fully redispersed samples fitted the master curve approach (surface charge vs. surface pH) reasonably well but still displayed an ionic strength dependence at a pH of less than 5 which could not be accounted for using the simple electrostatic model. Heterogeneity analysis of the master curve showed that the affinity distribution had two peaks centred at log KHint ∼ 4 and log KHint ∼ 8 to 9. The total number of weak acid sites titrated between pH 3.5 and 10.5 was approximately 3.5 eq kg−1 but the total number of sites estimated from the isotherm analysis was 5.3–5.8 eq kgt1¯. Double Toth and double Langmuir-Freundlich isotherms fitted the data almost equally well but the implied distribution of sites between the more acidic “car☐ylic” sites and the weakly acidic “phenolic” sites varied with the isotherm chosen. An important source of uncertainty in the analysis was in estimating the charge on the humic acid at its initial pH of about pH 3.

Analysis of ion binding on humic substances and the determination of intrinsic affinity distributions

Abstract Humic substances are characterized by a variable electric potential and by a variety of binding sites leading to chemical heterogeneity. Binding of ions to these substances is influenced by both factors. A methodology based on acid—base titrations at several salt levels is presented that allows for the assessment of an appropriate electrostatic double-layer model and the intrinsic proton affinity distribution. The double-layer model is used for the conversion of pH to pHS for each data point, where HS is the proton concentration in the diffuse layer near the binding site. It is shown that with an appropriate double-layer model the proton binding curves at different salt levels converge into one “master curve” when plotted as a function of pHS. The intrinsic proton affinity distribution can then be derived from the “master curve” using the LOGA method. A rigorous analysis of metal binding to humic substances is complex and in practice is not feasible. Under two different (simplifying) assumptions, namely fully coupled and uncoupled binding, it is shown how intrinsic metal ion affinity distributions can be obtained. Model calculations show that apparent metal ion affinity distributions do not resemble the intrinsic metal ion affinity distribution.

Humic matter and contaminants. General aspects and modeling metal ion binding

Humic substances are soil and fresh-water components that play an important role in the binding and transport of both organic and inorganic contaminants. Transport of the contaminants due to ground- and fresh-water dynamics is directly related to the risks associated with contaminations. The mobility of soluble humic substances is related to their interaction with soil mineral particles. Some key references for the binding of organic and inorganic contaminants and for the binding of humics to mineral particles are presented. Humic substances also play a role in the analysis of the contaminants in natural waters and with remediation of water or soil polluted with pesticides, heavy metal ions, and radionuclides. These aspects are illustrated with some examples. The problems that are encountered with the modeling of the binding of contaminants to humics and of heavy metal ions in particular are illustrated by considering the nonideal competitive adsorption model (NICA) extended with electrostatic interactions. The NICA-Donnan model gives quite good results for the description of metal ion binding, as is illustrated for metal ion binding to purified peat humic acid (PPHA). Finally, some remarks are made with respect to the use of the NICA-Donnan model in general purpose speciation programs and of simplified versions of the model for predictions under restricted environmental conditions.

Surface ionization and complexation models: A comparison of methods for determining model parameters

Abstract The triple-layer model for the electrical double layer is reviewed briefly with emphasis on the metal oxide-aqueous solution interface. To apply this method the model parameters must be established. Two general methods using acid-base titration data to evaluate the constants have been reported in the literature: a graphical double-extrapolation technique and a numerical optimalization of the adjustable parameters. Comparison of the results of both methods, as obtained for the TiO2 KNO3 system, shows a large discrepancy. To apply the graphical method the densities of both the positive and negative sites around the pzc should be small. This condition can only be satisfied if the difference between the two ionization constants, ΔpKi, is large. This assumption cannot be verified on the basis of titration data. A second assumption concerns the extent of complexation; this effect can be minimized by the double-extrapolation technique. For weakly adsorbing ions problems may arise with the determination of the complexation constants due to incomplete complexation. For strong specific adsorption of background electrolyte at the pzc a modification of the graphical method as suggested by Sprycha can be used, provided ΔpKi is large. In this case extra information regarding diffuse layer and specific adsorption is required. The numerical optimalization methods used are essentially nonlinear least-squares procedures for which the root-mean-square difference, Δ, between the measured and the calculated quantity is minimized. Calculations for the TiO2 KNO3 system show that it is not possible to find a unique set of parameters on the basis of the available titration data alone. However, a good fit of the experimental results can be obtained if either a small or a large ΔpKi value is chosen as the starting value for the calculation. The fact that no unique solution can be found is related to the large number of parameters to be optimized and the experimental and model inaccuracies. The latter determine the minimum value of Δ which is still relevant. It is concluded that neither the graphical nor the numerical method can give a satisfactory parameter set for the triple-layer model on the basis of titration data alone. Additional information regarding the magnitude of the ΔpKi value is required. Such information can be obtained from ψ0 (pH) measurements. Pseudo-Nernstian behavior of the electrodes occurs for small ΔpKi values, whereas strong deviations form Nernstian behavior occurs when the ΔpKi is large.