Chemical extraction and mobility

Chemical extraction is still the method of choice when assessing the environmental relevance of sediment-associated metals. It can give a first estimate of their mobility, provide us with information on changes of reactive sediment compounds (like Mn and Fe oxides, sulphides or organic matter), and elucidate the binding strength of metal-sediment associations. Although chemical extractions are still hampered by significant operational drawbacks, like re-adsorption and re-distribution phenomena or lack of selectivity, recent research confirms their successful applicability when characterizing or classifying metal-contaminated sites.

Chemical extraction is still the method of choice when assessing the environmental relevance of trace metals in sediments (or soils; see section 5.5 below). It rather easily allows to differentiate between potentially mobile metals and metals that strongly bind to the mineral lattice, or immobile. In particular, sequential extraction procedures (SEPs) are widely used, as they provide valuable informations about the species composition, bioavailability, pH and redox sensitivity of sediment-bound metals, that can not be obtained by total concentrations alone. They also allow us to predict possible changes in bioavailability if environmental conditions change.

However, and this is well accepted today, sequential extraction procedures (SEPs) are fraught with a high degree of uncertainty. At first, the obtained metal-binding phases are only operationally defined by the particular selected extraction procedure and may at best represent chemical species or not (cf section 5.4.2.4). Taking this into account, Dodd et al. (2000) studied contaminated anaerobic mud samples by means of 2 SEPs (Kersten and Forstner 1986, and Quevauviller 1998, cited in Dodd et al. 2000), and a Cryogenic Scanning Electron Microscopy (SEM) method coupled with Energy-Dispersive X-ray Analysis (EDXA), to check the extraction efficiency, reagent selectivity and possible reprecipitation. By analysing the unleached mud, the most abundant authigenic minerals were in decreasing order: Fe2+-phosphate (vivianite Fe3(PO4)2-8H2O) > mixed Fe-Cu-Zn sulphides > pyrite > calcite. Also, calcite did not completely dissolve during carbonate extraction, while vivianite began to dissolve in the carbonate extraction step of both SEPs, and was completely dissolved after the respective oxide extraction step. It was further shown that the ammonium oxalate residue still contained abundant Fe oxalate crystals, suggesting that

Fe from the dissolution of vivianite has been reprecipiated. The Fe oxalate dissolved in the subsequent sulphide extraction step. However, no sulphide or metal rich organic matter was found in the residue of the sulphide extraction step.

In this context, Dollar et al. (2001) used a Tessier-type sequential extraction scheme for metals in wetland sediments to study the metal remobilization potential, in particular with regard to possible implications of different restoration options. Their results gave a first indication, which metals may be affected by possible wetland reflooding. As Cr and Cu were mainly associated with oxidizable sediment fractions, these metals do not seem to be mobilised as a result of flooding, but may be even more stabilized. With flooding and corresponding wetter conditions the authors speculate that a concomitant decrease in redoxpotential may lead to the dissolution of oxides and hydroxides and the release of associated metals into the water. As Cd, Pb and Zn were mainly associated with Fe/Mn oxides, these metals may, consequently, be subject to remobilisation, if these sediments become flooded. Although only insignificant amounts present in the uppermost sediment, considerable amounts of Zn (beside Cd and Pb) may become remobilized at the same time from exchangeable fractions. By means of a simple mass balance calculation based on the performed chemical extraction, the authors assume that restoring only half of the most extensive part of the wetlands, the 'Great Marsh', with about 750 ha, would result in the total loss of the exchangeable Zn fraction (1125 ^g/cm2) in 1 year as a result of continuous flooding. From this it was estimated that about 84 t of Zn will be flushed away from the system, causing concentrations in the drainage water flowing into Lake Michigan as high as 5 ppm.

In the following some additional recent examples will be given to shortly underline the still existing difficulties and errors associated with the use of chemical extractions, which hamper a scientifically sound interpretation of the obtained data.

Ngiam and Lim (2001) used 3 different multi-step sequential extraction schemes to elucidate the speciation of metals in anoxic sediments. They found that sediment samples were already oxidized during the reducible fraction extracting step, resulting in an overrepresentation of metals associated with reducible phases and an underrepresentation of organic/sulfidic-bound phases, although more than 70 % of Zn (Cd and Pb) were associated with AVS.

When Lead et al 1998 studied the partitioning of metals into suspended particulates from a river, they observed a decrease of the metal binding capacity of the sample material when successively extracted by the Tessier method. However, transmission electron microscopy showed that the biological part of the suspended matter had undergone significant changes during extraction. This may have affected the original metal binding, and so complicated interpretation of the results with regard to simple mineral and organic phase reactions.

Gomez Ariza et al (2000) compared the feasibility of 3 well known sequential extraction schemes (i. e. the Tessier, Meguellati and BCR method) in sediments and provide an interesting discussion about their benefits and limitations. A more comprehensive discussion about the limitations associated with the use of chemical extraction schemes for metal speciation in aquatic sediments is given in Reuther (1999).

Also the fact that still a multitude of various digestion solutions of varying strength is used to assess the degree of total metal contamination, is contributing to handle current speciation data based on chemical extractants with great care. Sutherland et al. (2001) recently tried to test the capability of different extractants in identifying significant contaminant levels in road sediments. In their study, they included a total four-acid digestion scheme, a microwave-assisted digestion with conc. HNO3 (USEPA 3051), 0.5 M cold HCl, and 0.05M EDTA (pH 7). The authors summarized that the weak extractants they used (i. e. HCl and EDTA) were most successful to indicate the degree of anthropogenic metal contamination, why they recommend to use these agents more widely.

In a very recent comparison, Hseu et al. (2002) tested the applicability of different digestion methods to show, which method was best to assess total metal concentrations in soils and sediments. A great number of samples was subjected to aqua regia and different combinations of concentrated acids (HClO4, HNO3, H2SO4, HF). The results indicate that the 'Baker and Amacher' method (1982) (ref. given therein), consisting of a mixture of HF-HNO3-HQO4-H2SO4, performed best for Cd, Cr, Cu, Ni and Zn, and the 'Reisenauer' method (1982) for chromium, the latter using the same acid mixture but a modified heating process. The classical aqua regia gave also comparatively good results for Cu, Ni and Zn, while total Pb was extracted most efficiently by the so-called 'Burau' method (1982) (also cited therein), where the H2SO4 of the above acid mixture is replaced by HCl.

That chemical extraction, despite all these uncertainties, still provides a practical tool for getting valuable informations on the complex dynamics existing in sediments in regarding to the mobility, binding and bioavailability of metals, can be further underscored by the following examples.

According to van den Berg et al. (2001), assessing the mobility of contaminants is an important issue in the environmental risk assessment of soils and sediments. In particular, changes in the level of reactive phases (like Mn and Fe oxides, sulphides or organic matter), as derived by chemical extractions, may give us the necessary information on processes, which influence the behaviour of metals in sediments. The authors could demonstrate by means of a chemical extraction scheme that dissolved metal concentrations in water will not be significantly influenced by dredging operations, due to their obviously strong metal binding strength to solid phases, and/or their fast redistribution over sorptive phases in response to oxidation e. g. of metal sulphides.

There is a vast amount of literature now available, which can demonstrate the successful use of sequential extraction procedures to evaluate the efficiency of different sediment treatment techniques (see in particular recent publications in Environmental Science and Technology). When used before, during and after a particular treatment, changes in metal extractability/leachability or possible species transformations are indicated. But it becomes also obvious, from which sediment pools (e. g. carbonates, Fe and Mn oxides, organic matter and sulphides) metals are released (see for more details also section 5.5.4) (Chen et al. 2000; Chartier et al 2001).

Clark et al. (2000) argue that size normalization during sequential chemical extraction may be a rather safe and easy way to directly compare sediments from dissimilar environments, and may in particular help to answer the question, whether a sediment is a source or sink for trace metals.

In order to integrate Microtox toxicity data (EC50 values), sediment contaminant concentrations and speciation data from sequential extraction, Mowat and Bundy (2002) developed a mathematical algorithm. They calculated a 'toxicity index' (TI) as an indicator of adverse ecological effects, which can be used to rank metal contaminants according to their particular toxicity. Summarizing, the authors conclude that concentrations using bioavailability data from sequential extraction were found to be the best theoretical predictors of the observed experimental mixture toxicity value.

How helpful chemical extraction can be used, when assessing the toxicity of surface sediments, may be reflected by the following short example. Borgmann and Norwood (2002) observed that sediment profiles of the total metal concentration from a lake near Sudbury, Ontario (Canada), differed from corresponding bioavailability profiles, as derived from chemical extraction. Likewise, Cu bioavailability profiles differed from profiles obtained for Cd, Co and Ni. In fact, deepest preindustrial sediment layers proved to be non-toxic, while the observed sediment toxicity at the surface could be attributed to the dissolution of Ni from labile phases into the overlying water. Chemical extraction so could help to trace the occurrence of different metal bioavailabilities in surface and deeper sediment layers, a fact, which has to be considered when interpretating and comparing toxicity data for sediments taken by different sampling techniques (e. g. by grab or core samplers).

0 0

Post a comment