Twenty Dutch soils were sampled at sites that were expected to have elevated levels of heavy metals compared to background values. Field‐based partition coefficients (Kp) for As, Cd, Cr, Cu, Ni, Pb, and Zn were determined by calculating the ratio of the amount of metal extracted by concentrated HNO3 to the metal concentration in the pore water. Kp values varied widely with metal and soil type and correlated well with soil pH. To a lesser extent amorphous Fe‐content and dissolved organic carbon (DOC) also explained part of the variation in Kp values. Thereupon, regression equations based on easily obtainable soil characteristics were derived. External validation of the regression equations showed that predicted log Kp values were in agreement with measured values.

Abstract-To evaluate the adequacy of the equilibrium partitioning concept in predicting metal bioaccumulation, a soil invertebrate species was exposed in 20 Dutch field soils with moderate metal contamination. Earthworms (Eisenia andrei) were kept in the soils for 3 weeks under laboratory conditions. Bioconcentration factors (BCFs) for six metals (Zn, Cu, Pb, Cd, Cr, Ni) and for As were calculated as the ratio of body-and solid-phase metal concentrations. Multivariate statistical analyses suggested that the BCFs for As, Cd, Cu, and Zn are governed by the same soil characteristics that determine equilibrium partition coefficients between the soil solid phase and the pore water. This suggests that uptake of metals is either direct from the pore water or indirect through an uptake route closely related to pore water. Regression equations were derived for predicting BCF values as a function of easily determinable soil characteristics. By means of internal validation it was shown that the equations obtained can be used for predictive purposes within the range of soil properties encountered in the dataset. Due to a lack of data, external validation was possible only in a qualitative sense.

Abstract-Twenty Dutch soils were sampled at sites that were expected to have elevated levels of heavy metals compared to background values. Field-based partition coefficients (K p ) for As, Cd, Cr, Cu, Ni, Pb, and Zn were determined by calculating the ratio of the amount of metal extracted by concentrated HNO 3 to the metal concentration in the pore water. K p values varied widely with metal and soil type and correlated well with soil pH. To a lesser extent amorphous Fe-content and dissolved organic carbon (DOC) also explained part of the variation in K p values. Thereupon, regression equations based on easily obtainable soil characteristics were derived. External validation of the regression equations showed that predicted log K p values were in agreement with measured values.

Summary Al hydroxide polymers (AlHO) can significantly influence the cation exchange behaviour of clays. We have determined the effect of synthesized AlHO on Ca–Na, Zn–Na and Pb–Na exchange for a series of exchanger compositions and two Al loadings at pH 6.0 and an ionic strength of 0.01 m. The preference for Ca on the siloxane surface of the clay–AlHO system (CAlHO) was greater than for the pure clay, and the average KV (Vanselow selectivity coefficient) was determined to be 2.16 and 1.24, respectively. The selectivity coefficients for the exchange reactions Zn–Na and Pb–Na were not directly determined in CAlHO systems, because heavy‐metal ions bind as well to the clay surface as to the AlHO over a wide range of pH. We have estimated the effect of the presence of AlHO on the selectivity coefficients of Zn–Na and Pb–Na exchange by extrapolation of the experimental results of Ca–Na, Zn–Na and Pb–Na exchange for pure clay and Ca–Na exchange for CAlHO. The average KV was increased by the presence of the AlHO from 1.23 to 2.16 for Zn–Na exchange and from 1.59 to 2.77 for Pb–Na exchange. The increase in the preference for the divalent cations is probably caused by parallel alignment of clay platelets by sorption of AlHO. Increasing the amount of AlHO did not change the selectivity for Ca–Na exchange, and probably the structure of the system or the arrangement of the clay platelets and AlHO particles was not substantially changed. This was supported by the linear reduction of the cation exchange capacity with amount of AlHO present at pH 6.6. It seems likely that the selectivity coefficients for Ca–Na, Zn–Na and Pb–Na exchange that we found apply in naturally occurring montmorillonite–AlHO systems.

Groundwater samples were taken from seven bore holes at depths ranging from 2 to 41m nearby drinking water pumping station Vierlingsbeek, The Netherlands and analysed for Y, La, Ce, Pr, Nd, Sm and Eu. Shale-normalized patterns were generally flat and showed that the observed rare earth elements (REE) were probably of natural origin. In the shallow groundwaters the REEs were light REE (LREE) enriched, probably caused by binding of LREEs to colloids. To improve understanding of the behaviour of the REE, two approaches were used: calculations of the speciation and a statistical approach. For the speciation calculations, complexation and precipitation reactions including inorganic and dissolved organic carbon (DOC) compounds, were taken into account. The REE speciation showed REE(3+), REE(SO(4))(+), REE(CO(3))(+) and REE(DOC) being the major species. Dissolution of pure REE precipitates and REE-enriched solid phases did not account for the observed REEs in groundwater. Regulation of REE concentrations by adsorption-desorption processes to Fe(III)(OH)(3) and Al(OH)(3) minerals, which were calculated to be present in nearly all groundwaters, is a probable explanation. The statistical approach (multiple linear regression) showed that pH is by far the most significant groundwater characteristic which contributes to the variation in REE concentrations. Also DOC, SO(4), Fe and Al contributed significantly, although to a much lesser extent, to the variation in REE concentrations. This is in line with the calculated REE-species in solution and REE-adsorption to iron and aluminium (hydr)oxides. Regression equations including only pH, were derived to predict REE concentrations in groundwater. External validation showed that these regression equations were reasonably successful to predict REE concentrations of groundwater of another drinking water pumping station in quite different region of The Netherlands.