Return to search

Using multi-element stable isotope dilution to quantify metal reactivity in soil

Determining the total concentration of elements in soils seldom provides sufficient insight into trace metal bioavailability. However, measurement of ‘isotopically exchangeable’ metal can provide a better evaluation of metal reactivity and potential toxicity. Traditionally this requires the use of problematic radio-isotopes (e.g. 109Cd (γ)). Fortunately, increasing access to Inductively Coupled Plasma Mass Spectrometry (ICP-MS) in recent years has led to greater use of enriched stable isotopes of trace metals. The lability of heavy metals has been determined through a variety of approaches, including single and sequential extraction or predicted by geochemical models. In the present work, multi-element stable isotopes methods were developed for simultaneously determination of the labile pool of Fe, Zn, Cd and Pb using isotopic exchange principles. This included experimental and instrumental development for an accurate and precise determination of labile metal pool in soils. This approach was then validated by quantifying Zn, Cd and Pb in contaminated soils (Derbyshire; n = 8 and Weardale catchment; n = 246) and comparing the outcome results with common traditional extraction procedures. The variation of metal lability with soil characteristics was used to predict metal lability from the simple soil measurements using a multiple regression approach. In addition, E-values of Fe, Zn, Cd and Pb was used as input to WHAM(VI) (Windermere Humic-Aqueous Model) to predict metal solubility, emphasising in the role of Fe under reducing conditions in this regard. The results showed that isotopic dilution is a robust mechanistic method for assessing the ‘reactive’ pool of multiple trace metals over a wide range of soil characteristics. The results showed a very wide range of metal reactivities (almost 1%-100%) for Zn, Cd and Pb that were consistent over a range of spike concentrations. Sub-micron forms of non labile metal are perhaps most likely to occur in suspension either strongly bonded to humic/fulvic acids or occluded within CaCO3 particles. It appears that E values have no consistent correspondence to any chemical extraction procedure. Nevertheless, the use of 0.43 M HNO3 to extract labile metal in organic soils at pH < 6 appears justifiable - especially where humus is likely to be the principal adsorption surface. It is also important to acknowledge that extractions are not necessarily intended to estimate the entire reactive fraction. Thus, DTPA has been successfully applied as an empirical prediction of plant uptake but its extraction capacity is particularly limited in calcareous systems where it substantially underestimates the isotopically exchangeable metal pool. Speciation calculations showed that prediction of metal solubility was much better when the isotopically reactive metal pools were used as input to WHAM(VI). The soil samples that fitted best had pH values less than 4.0 and high organic matter contents reflecting the strength of the humic binding component of WHAM(VI) particularly in the case of Zn. The changes in metal solubility and lability under reducing conditions were mainly affected by pH. Moreover, the measurement of Fe2+ in the solution phase was considerably lower than that of the isotopically labile Fe2+ which calls into question the dependence on soluble Fe2+ to predict reductive dissolution of Fe-oxides. In addition, under reducing conditions the variables input of Fe to WHAM(VI) showed greatest effects on predicting metal solubility. It was found that Zn and Cd were affected only by Fe2+ competition for adsorption sites while predicted Pb solubility was more affected by loss of oxides than competition processes. The fractionation results, output from WHAM(VI), showed that a significant proportion of Pb was associated with Mn-oxides. Therefore, the calculation of loss of the adsorption site of Mn-oxides depending on Mn2+ measured in the solution phase did not improve the predicted Pb solubility where the model underestimate the adsorbed labile Mn as inference from Fe results.

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:587843
Date January 2012
CreatorsMarzouk, Ezzat Rashad El-Said
PublisherUniversity of Nottingham
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttp://eprints.nottingham.ac.uk/28914/

Page generated in 0.0022 seconds