New spectroscopic approaches

Spectroscopic approaches are increasingly used today as direct means to assess environmentally relevant metal species in particulate matters. In particular, X-ray absorption spectroscopic techniques (XAS) have almost revolutionized our present understanding on metal speciation, as they identify and measure discrete metal or ligand species of solid surfaces (of minerals or cells), as well as surface characteristics, which may be critical. By providing the necessary structural information, these techniques present a robust molecular basis for method interpretation and verification, and allow improved understanding of results obtained from other speciation methods (like AVS or chemical extraction). By some of the new techniques, it is possible to distinguish the dominant metal-ligand species in sediment samples, elucidate if metals are really present in the form (e. g. as sulphide or oxide, or as a mixture of both) indicated, e. g. by chemical extraction, and hence establish the true species composition.

New techniques such as Transmission Electron Microscopy (TEM) coupled with Energy Dispersive Spectroscopy (EDS) have been successfully applied in speciation studies to provide information on particle size, morphology, crystallinity and composition and have revealed that sediment particles are associated with biological cells and even surrounded by extracellular structures. It is supposed that these particles act as important metal carriers in the aquatic environment, and as catalyst sites for further particle formation. Obtained results emphasize also the need to work at the micro-scale to identify relevant contaminant-bearing phases in aquatic ecosystems, and to help to understand the role of biological surfaces, as important sites for metal speciation.

On the one hand, there is a need for rapid and routine sediment contamination and toxicity assessments, which underscores recent efforts by various controlling authorities to establish more standardized testing methods and quality criteria for metals in sediments. However, these test methods and criteria have to be put on a scientifically sound basis. For this reason, there is also a growing demand on the other hand, within the regulatory community, to develop more accurate sediment quality protocols for trace metals. Both can be achieved, for example, by coupling conventional chemical speciation methods more closely to standardized toxicity tests, and more recently, to spectroscopic analysis. This type of more synergistic biological-geochemical approach may successively improve already existing attempts to assess toxicity as a function of metal concentration, sediment type, porewater geochemistry, and may reflect more accurately the true range of geochemical processes, for example responsible for removal or uptake of metals from the porewater. As a matter of fact, new spectroscopic techniques are increasingly considered as a direct means to assess relevant metal species, as they represent a robust molecular basis for method interpretation, verification and for the transferability of results obtained from more conventional chemical and biological tests, among a range of aquatic environments.

In particular, absorption spectroscopic approaches have almost revolutionized our present understanding of metal speciation phenomena, as they allow us to directly identify and measure different metal or ligand species in solid materials (like minerals and cells) as well as surface charcteristics, which may be critical. This new class of speciation techniques also helps to validate the role of new concepts, like the SEM/AVS ratio in soils and aquatic sediments, or of the more commonly used chemical extraction procedures (see section before). In the following, some typical examples will be presented to illustrate how spectroscopic methods are successfully applied now to elucidate the complex nature of solid metal speciation.

O'Day et al. (2000) tried to combine the SEM/AVS method with invertebrate toxicity testing and X-ray absorption spectroscopy (XAS), to evaluate metal speciation and the ecological hazard of estuarine sediments from the Seaplane Lagoon, San Francisco Bay. This site that has been contaminated by military and industrial activities, but should now be returned to the public and private sector. Bay sediments are characterized by a moderate to low toxicity and by increasing sediment and porewater metal concentrations with depth. SEM/AVS ratios and results from toxicity tests were compared with the molecular binding of metals in sediment solid phases measured by 'X-ray Absorption Near-Edge Structure' (XANES) and 'Extended X-ray Absorption Fine Structure' (EXAFS) analysis, to show whether metals are present as sulphides or oxides, or as a mixture of both. The results confirm that some assumptions behind the SEM/AVS approach may not be fully valid:

1. Of all the metals studied in the sediment, only Cd was exclusively present as sulphide (namely as amourphous or poorly crystalline

2. About 80% of Zn and 50-70% of Mn were present as sulphide phase (most probably as sphalerite ZnS and alabandite MnS), and the rest remained associated with oxides.

3. All of Cr and Pb in the sediment was associated with oxide phases.

4. Fe occurred primarily as Fe(II) in pyrite and clay minerals; and

5. none of the metals investigated was coordinated by organic ligands.

Table 5.8. Local structure of Zn in reference compounds, determined by multishell fit of Zn K-edge EXAFS analysis (from Scheinost et al., 2002)

first shell second and third shell

Table 5.8. Local structure of Zn in reference compounds, determined by multishell fit of Zn K-edge EXAFS analysis (from Scheinost et al., 2002)

first shell second and third shell

Reference mineral

formula

CNa and

R[Â]b

62[Â2]

CN and

R[Â]

62

element

c

element

[Â2]

willemite

Zn2SiO4

4.8 O

1.95

0.005

2.9 Zn

3.25

0.009

hemimorphite

4.1 O

1.94

0.006

21.4 Zn

3.33

0.028

Zn4Si2O7(OH)2H2O

ZnAl2 O4

4.4 O

1.97

0.005

14.5 Al

3.41

0.005

gahnite

Franklinite

4.0 O

1.96

0.003

12.0 Fe

3.51

0.007

(Zn,Fe,Mn)"(Fe,Mn)In2O4

3.22 O

2.33

0.004

9.0 Zn

3.81

0.009

sphalerite

ZnS

9.1 s

4.46

0.01

aqueous

Zn2+

5.7 O

2.07

0.010

Zn oxalate

ZnC2O4-2H2O

n

4.3 C

2.80

0.005

lithiophorite

6.4 O

2.02

0.009

4.7 Al

2.95

0.004

(Li,Al)(Mn)O2(OH)2

Zn-sorbed soil

6.9 O

2.08

0.008

0.6 Al

2.99

0.001

Zn-sorbed HIM

5.7 O

2.05

0.010

3.8 Al

3.05

0.006

Zn-copr. HIM

6.3 O

2.07

0.009

6.0 Al

3.06

0.006

Zn-Al LDH

6.3

2.07

0.009

3.9 Zn

3.10

0.008

[Zni-xAlx(OH)2]+(x)Cl-mH2O

6.5 O

2.01

0.013

6.3 Mn

3.42

0.007

chalcophanite (Zn,Fe,Mn)nMn3IVO7-3H2O

a Coordination number, bRadial distance, cDebye-Waller factor chalcophanite (Zn,Fe,Mn)nMn3IVO7-3H2O

a Coordination number, bRadial distance, cDebye-Waller factor

Surprisingly, pyrite was the only Fe sulphide present in the sediment. Concentrations of Cu and Ni were too low to allow good X-ray absorption spectra to be obtained. It was demonstrated that HCl-extraction dissolved also other sedimentary metal phases. Although Pb dissolved rapidly after 24 h in HCl, it was not present as sulphide, as evidenced by EXAFS and a geochemical modelling of the porewater suggested its association with carbonates (very soluble in strong acid) or sorption as an oxygen-ligated surface complex (also desorbing in strong acid). Like Cu and Ni, also Cr was poorly extracted in HCl. XAS analysis again indicated association with oxides. Table 5.8 just illustrates the type of structural information we obtain by XAFS spectroscopy necessary to determine the most dominant species for Zn (here reference compounds), as an example, in a sediment (or soil) sample.

In a similar study, also Cooper & Morse 1998 (see also section 5.4.3.3) showed that Cu and Ni sulphides (e. g. NiS, NiS2, Ni3S2, CuS, Cu2S) are poorly extracted by cold acid (1-31 % after 1 h in 6 M HCl). Without spectroscopic evidence, it remains uncertain whether Ni and Cu are present as sulphides in the sediment. Like Cr, also Ni seems most likely to be associated with weathering products (such as secondary clay minerals).

The vertical distribution of AVS in the sediment showed a significant concentration increase with depth. However, the performance of a dissolution time series uncovered that at 30 cm depth all AVS dissolved already after 30 min, whereas metal dissolution was not complete at the same time, even after 24 h. Measuring sediment charcteristics showed that dissolved O2 was removed within the first 2-10 mm below the sediment surface, and that dissolved sulphide below a depth of 30-40 mm occurred throughout the sediment core. Based on porewater concentrations the authors supposed that dissolved sulphide (HS-) accounts for most if not all AVS extracted from the sediment. Also, the obtained spectroscopic data could not confirm that FeS(s) is the primary contributor to AVS in these sediments as commonly assumed in the ^SEM/AVS measurement (O'Day et al. 2000).

The observation that metals in reduced anoxic sediments may occur both as sulphide and oxide solid phases is still rather controversial among experts and has implications for establishing sediment quality values (see chapter 8). The study of O'Day et al. (2000) at least shows that the XSEM/AVS concept was indeed only valid for Cd and to some extent for Zn. Consequently, oxidative dissolution of metal sulphides would only affect these two metals. Other metals, like Cr and Pb, and also Ni and Cu, seem to bind more to oxide and silicate surfaces, maybe as inner-sphere complexes and precipitates of carbonate and oxyhydroxide, which are pH-dependent. Partitioning of these metals seems more affected by early diagenetic changes in pH and mineral substrate stability (aging!) occuring within the uppermost part of the sediment profile (see also section 5.5.6).

When it comes to sediment toxicity, the ^SEM/AVS ratios of 2.75.25, measured by O'Day et al (2000) did not support the toxicity prediction they suggest, mainly because porewater metal concentrations at the sediment-water interface were too low. This was in contrast to deeper sediment layers, where a 100% toxicity was observed at ^SEM/AVS = 0.54, probably due to high AVS (sulphide!) values. However, further porewater analyses showed that the observed invertebrate toxicity possibly was attributable to ammonia and low oxygen levels, although this was not easy to verify due to variable organism responses and the geochemical heterogeneity of the sediment. Based on the toxicity and spectroscopic data, the authors conclude that the ^SEM/AVS threshold (<1) is not a good criterion for predicting acute organism toxicity in aquatic sediments, a fact which has been already acknowledged by the US EPA in the mean time (see ref. therein). In particular, the obtained spectroscopic results indicate that the equilibrium partitioning given by Me2+(aq) + FeS(s) = MeS(s) + Fe2+(aq) as rationale for the ^SEM/AVS concept may be valid only for Cd, partially for Zn, but not for Pb, Cu and Ni in these estuary sediments. So, at its best measuring AVS may be a useful indicator for Cd and Zn availability in reduced sediments, but not a general measure of bioavailability for trace metals in a variety of geochemcial environments as proposed.

Organic and inorganic particles (e. g. humics, polysaccacharides, microorganisms, Fe and Mn oxides/hydroxides, clays) among others are supposed to mediate the cycling of essential and toxic metals in aquatic systems. 'Transmission Electron Microscopy' (TEM) coupled with 'Energy Dispersive Spectroscopy' (EDS) has been successfully applied before in speciation studies to provide informations on size, morphology, crystallinity and composition of sediment particles, and on possible physical and biological associations. Previous artefacts produced through the dehydration process during sample preparation leading to shrinkage and aggregation of particles can now be overcome by utilizing a hydrophilic resin ('Nanoplast') to embed the sample. Web et al. (2000) examined the direct association of zinc in a highly contaminated lake sediment with up to 30 % Zn d.w. at the individual particle scale by TEM/EDS analysis, and the influence of biological processes on Zn distribution.

Electron microscopy revealed the existence of different colloidal particles in more oxic lake sediments, including bacteria, small algae, organic fibrils and matrices, clay minerals, Fe hydroxides, Fe-Zn-phosphates, biogenic silica and diatoms, sulfidic minerals, and to a minor extent carbonates of Ca and Mg. Zn-bearing colloids were often strongly associated with or entrained into biological structures. They seem intimately associated with algal cells, whose organic-rich outer surface may act as a nucleation site for particle formation. In these particles Zn may coprecipitate or coentrain at the observed high concentration, and not only become bound by surface adsorption. Other Zn-containing particles included Fe-Zn-oxihydroxides and to a lesser extent Zn sorbed to Fe-rich clay minerals. Also particles in the water column (with up to 300 ^M Zn) consisted mainly of diatom frustules, clay minerals and organic matrices. In general, Zn was at most associated with Fe-P-rich particles or sorbed to Fe/Mn-rich precipitates on bacterial cell walls, but not on clay minerals, demonstrating the clear relationship between biological structures and Zn partitioning. In contrast, in a more anoxic lake sediment sample, particulate Zn was associated with small rounded globules (45-210 nm size) and mineralized layers, together with high sulfur concentrations suggesting a diagenetic formation of ZnS.

Speculating on the origin of the observed Fe-P-Zn-bearing particles in the sediment, Webb et al. (2000) resume that they may be caused by biological processes leading to phosphorus precipitation on the cell wall. Indeed, TEM/EDS anaylsis showed that these particles are associated with biological cells and surrounded by extracellular structures, and may act as Zn carriers in the lake, but also as catalyst site for particle formation. The authors claim more research to elucidate processes occuring at the cell wall level and leading to the formation of colloidal metal-bearing particles. They also refer their results to the current literature, from which we know that fibrils (extracellular polymers) on bacterial surfaces provide appropriate nucleation sites for Fe and Mn oxides, that bacterial cell walls can remove and accumulate metals from dilute solutions, that generation of mineralized scales (Ca, Si) and wall parts of colloidal dimensions induced by internal metabolic processes in microalgae occur, and that Fe oxyhydroxide coatings on cell walls and extracellular fibrils strongly can accumulate Cu from porewater. All these studies prove the importance of the microscale to identify relevant contaminant-bearing phases in aquatic ecosystems, and of the biological templates as important sites for contaminant transformation. However, one short-cut of the used approach was the used X-ray detector, which was only sensitive to relatively high elemental concentrations (> 0.1 %) and hence does not cover particles with smaller metal concentrations, but sufficiently abundant to make up a significant fraction of the total metal concentration. Complementary techniques, like XAS would help to overcome this limitation (more speciation studies using spectroscopic analysis are presented in sections 5.5.4, 5.5.5, and 5.5.6, below).

0 0

Post a comment