Predicting trace metal solubility and fractionation in Urban soils from isotopic exchangeability

Abstract

Metal-salt amended soils (MA, n = 23), and historically-contaminated urban soils from two English cities (Urban, n = 50), were investigated to assess the effects of soil properties and contaminant source on metal lability and solubility. A stable isotope dilution method, with and without a resin purification step, was used to measure the lability of Cd, Cu, Ni, Pb and Zn. For all five metals in MA soils, lability (%E-values) could be reasonably well predicted from soil pH value with a simple logistic equation. However, there was evidence of continuing time-dependent fixation of Cd and Zn in the MA soils, following more than a decade of storage under air-dried conditions, mainly in high pH soils. All five metals in MA soils remained much more labile than in Urban soils, strongly indicating an effect of contaminant source on metal lability in the latter. Metal solubility was predicted for both sets of soil by the geochemical speciation model WHAM-VII, using E-value as an input variable. For soils with low metal solution concentrations, over-estimation of Cd, Ni and Zn solubility was associated with binding to the Fe oxide fraction while accurate prediction of Cu solubility was dependent on humic acid content. Lead solubility was most poorly described, especially in the Urban soils. Generally, slightly poorer estimation of metal solubility was observed in Urban soils, possibly due to a greater incidence of high pH values. The use of isotopically exchangeable metal to predict solubility is appropriate both for historically contaminated soils and where amendment with soluble forms of metal is used, as in toxicological trials. However, the major limitation to predicting solubility may lie with the accuracy of model input variables such as humic acid and Fe oxide contents where there is often a reliance on relatively crude analytical estimations of these variables. Trace metal reactivity in urban soils depends on both soil properties and the original source material; the WHAM geochemical model predicts solubility using isotopically exchangeable metal as an input.

The labile fractions of metals and arsenic in mining‐impacted soils are explained by soil properties and metal source characteristics

“…The dashed lines are previously established relationships of labile fractions in well‐equilibrated metal salt amended soils derived from Mao et al. (2017).…”

“…Such models have been taken forward to regulations (Smolders et al., 2009). Selected empirical trend lines are added in Figure 3 (Mao et al., 2017). In general, these lines are on the higher end of labile fractions observed here, indicating that the labile fractions in the mining‐impacted soils are lower than in the metal salt‐impacted soils and that some nonlabile metal may originate from the ore lithology.…”

“…Marzouk, Chenery, & Young, 2013b;Smolders et al, 2012). Multivariate models were fitted to data of %E of Cd, Cu, Ni, Pb, and Zn of a set of 3-yr aged metal salt-amended soils (n = 23) and a second set of urban contaminated soils (n = 50) (Mao, Young, Tye, & Bailey, 2017). In general, the contaminated soil samples exhibited similar trends between %E and soil properties as in metal salt spiked soils, but %E values were lower than those in the metal salt-amended soils, illustrating the role of the contamination source on metal lability.…”

“…Labile fractions of Ni and Pb were very comparable with the same set of reported values. A main reason for the discrepancies is the type of sampled analyzed, as T A B L E 3 Significant (p < .01) linear regression parameters (m 1 to m 4, Equation 4) from stepwise multivariate analysis of labile fractions (%E) of Ni, Cu, Zn, Cd, Pb, and As as a function of soil chemical properties and contamination source attributes Mao et al (2017). in the case of Buekers et al (2007), where high fractions of labile Cu and Cd correspond to control-uncontaminated soils. When field-contaminated soils are analyzed (as in Garforth et al (2016) and Marzouk et al (2013b), the discrepancies are more related to soil chemical properties (mainly to lower pH values and oxalate-extractable Fe contents compared with the values of the present work).…”

“…Surveys of environmentally contaminated soils have shown that the role of soil properties, such as pH, relate only poorly to the %E values of Zn, Cd, Pb, and Cu and that %E is generally below that of the pH‐dependent labile fractions in soils that had been spiked with metal salts (Degryse, Buekers, & Smolders, 2004; Marzouk, Chenery, & Young, 2013b; Smolders et al., 2012). Multivariate models were fitted to data of %E of Cd, Cu, Ni, Pb, and Zn of a set of 3‐yr aged metal salt–amended soils ( n = 23) and a second set of urban contaminated soils ( n = 50) (Mao, Young, Tye, & Bailey, 2017). In general, the contaminated soil samples exhibited similar trends between %E and soil properties as in metal salt spiked soils, but %E values were lower than those in the metal salt–amended soils, illustrating the role of the contamination source on metal lability.…”