Combining trace elements micro-analysis in deposited dredged

EMAS 2011 WorkshopOPEN ACCESS

View the article online for updates and enhancements.
-
-
-
-
This content was downloaded from IP address 114.25.24.34 on 12/10/2021 at 20:45
A Poitevin1,4, C Lerouge1, G Wille1, P Bataillard1, P. Quinn2 and L Hennet3 1 BRGM, 3 avenue Claude Guillemin, P.O. Box 36009, FR-45060 Orleans cedex 2, France 2 Diamond Light Source Ltd., Didcot OX11 0DE, Great Britain 3 CEMHTI, CNRS, FR-45071 Orleans cedex 2, France E-mail: [email protected] Abstract. Since deposited dredged sediments are rich in metallic contaminants, they present a risk for environment. This work aims to study dredged sediments chemical composition, identify metal-carrier minerals and understand their mobility. Combining chemical and spectroscopic techniques at multi-scale for an integrative approach of trace elements (zinc, lead, iron) behaviour is therefore necessary. The global mineralogy and the chemistry of the sediment were determined by X-ray diffraction and fluorescence (XRF), respectively. Zn and Pb enriched fractions were separated using a sequential chemical extraction procedure and measured by inductively coupled plasma atomic emission and mass spectroscopy. Microanalyses using scanning electron microscopy (SEM), electron microprobe microanalysis (EPMA), combined with synchrotron radiation X-ray fluorescence (µ-XRF) were carried out to characterize mineralogical phases and identify Zn and Pb carrier minerals. Iron oxyhydroxides and iron sulphides were consistently identify as Zn and Pb carriers. The assumption that carbonate fraction was the major Zn carried phase, as demonstrated by chemical extraction results, was not verified by EPMA or µ-XRF.
1. Introduction Sediments originating from periodic dredging of waterways were traditionally deposited on soil without specific precaution or planed treatments. It may be of environmental concern especially when they came from areas historically contaminated with trace elements (e.g., Zn, Pb) like in the north French coal basin and considering the volume of sediments dredged annually (~ 200000 m3). The chemical risk is the breakdown of trace metals-carrier phases (electronic structure: oxidation state and nature of chemical bonds; crystal chemistry) liberating toxic elements which can migrate to the underlying soil and groundwater. In this framework, dredged sediments are currently studied to determine trace element contents, their distribution among phases, and the stability of these phases.
The study site is a ten years old dredged sediment deposit from the Canal de Lens (northern France) located in the district of Courrières (62). The whole amount of contaminated sediments deposited on soil reaches 3 metres. The main purpose of this work is to better understand and constrain mobility and bioavailability of trace elements along an annual cycle, more particularly to identify and characterize the Pb- and Zn-carrier minerals by combining chemical and spectroscopic techniques. Sequential chemical methods [1] have been traditionally applied to various contaminated 4 To whom any correspondence should be addressed.
EMAS 2011: 12th European Workshop on Modern Developments in Microbeam Analysis IOP Publishing IOP Conf. Series: Materials Science and Engineering 32 (2012) 012021 doi:10.1088/1757-899X/32/1/012021
Published under licence by IOP Publishing Ltd 1
soils [2, 3]. It is now widely recognized that this indirect and operational approach should be combined with other physical techniques, which allow direct identification of metal forms, such as scanning electron microscopy (SEM), electron probe microanalysis (EPMA), and synchrotron radiation techniques such as X-ray absorption spectroscopy (XAS), diffraction (XRD) and fluorescence (XRF) [4-7].
In this framework, the study presented here integrates three approaches: (i) traditional chemical sequential extraction method; (ii) reconstruction of the mineralogy of the sediment combining X-ray diffraction (XRD) and chemical analyses of major elements by X-ray fluorescence (XRF) of bulk sediment; (iii) identification of Zn and Pb-carrier minerals using optical and SEM observations, followed by quantification of Zn and Pb contents using EPMA and μ-XRF. 2. Materials and methods 2.1. Sample collection and preparation Samples were collected at the deposit surface (0 - 20 cm). They were conditioned in closed plastic bag with limited air at 4 °C on the field then in a fridge at the laboratory. They were homogenized and divided into two parts. The first one was directly used for chemical sequential extractions. The other one was dried at 105 °C during 48 h. This dried sediment was for one part used for chemical bulk analyses (X-ray, ICP-AES, XRD), and for a second part embedded in a polyacrylamide resin (LR White®) and mounted into polished thin sections for microanalysis (SEM, EPMA, µ-XRF). 2.2. Scanning electron microscopy Observations, analyses and elemental mapping were carried out with a Jeol JSM 6100 scanning electron microscope coupled with an energy-dispersive X-ray spectrometer (Kevex Quantum) using an acceleration voltage of 25 kV. Prior to the analysis, samples were coated with a 20 nm carbon conductive layer. 2.3. Electron probe microanalysis EPMA analyses and elemental mapping were carried out on carbon-coated polished thin sections using a Cameca SX-50 electron microprobe. Elemental mapping of major elements (Si, Al, Ca, Fe) were performed with a 15 kV acceleration voltage and a beam current of 15 nA. EPMA analyses of Ca, Zn and Pb on carbonate grains were carried out under analytical conditions of 25 kV and 50 nA. Ca, Zn and Pb peak intensities were measured with counting times of 240 s for punctual analyses and about 600 s for analyses along transects to improve detection limits. Elemental analyses were performed by employing 4 spectrometers. Ca Kα and Pb Mα were analyzed using a PET (pentaerythritol) crystal, and Zn Kα was analyzed using a LiF (lithium fluoride) crystal. Standards used included both well characterized natural minerals and synthetic ones: apatite (CaPO4) for Ca, sphalerite (ZnS) for Zn, and galena (PbS) for Pb. In these conditions, detection limits, calculated from [9], are in carbonate about 175 mg/kg for Ca and 125 mg/kg for Zn. Pb was not detected. The PAP matrix correction model was used for quantitative microanalysis [8]. Background subtractions were determined for each analysis by taking into account potential lines interferences. 2.4. X-ray diffraction The nature of crystalline phases present in the samples was determined by XRD using a Siemens D5000 diffractometer working in a Bragg-Bentano geometry and equipped with a scintillation counter. Diffraction patterns were recorded using Co Kα1 radiation (40 kV, 30 mA) over a 2θ angular range of 4 to 84° for powders and 2 to 36° for clayey fractions. In both cases, 2θ step of 0.02° and counting time of 1 s were applied. X-ray diagrams were analyzed using Diffrac-Plus software in an unquantitative way.
2
2.5. X-ray and µ-X-ray fluorescence The global chemical composition was determined by XRF using a wavelength-dispersive X-ray spectrometer Panalytical PW2400. Prior to the measurements, powders were heated at 1000 °C during 2 hours to measure the loss of ignition, then mixed with lithium bromide and lithium tetraborate and then heated at 1200 °C to vitrify samples. Concentrations were calculated from a calibration line of synthesized or natural standard concentrations. Detection limits varied in the range of 200 mg/kg (manganese oxide MnO) to 2000 mg/kg (for alumina Al2O3 and silica SiO2).
µ-XRF experiments were carried out at the I18 microfocus beamline at Diamond Light Source (Didcot, UK). The beam size was about 3 µm x 3 µm and the incident energy was 16.5 keV. A 9-elements germanium solid state detector placed perpendicularly to the beam was used to measure the fluorescence signal from the sample. Elemental mappings on polished thin sections were performed in the 4 - 16.5 keV energy range. The data treatment was carried out using the PyMCA software. 2.6. Sequential chemical extractions protocol Chemical analysis was performed onto three replicates of each dried bulk and wet surface sediments. A 5-step extraction procedure was applied on 0.5 g of wet deposited sediment according to the method described by Piou et al. [8] and summarized in table 1.
Table 1. Sequential chemical extraction procedure adapted from Piou et al. [10].
Fraction Procedure Mineralogical extracted compartment
F1 : Leachable Calcium nitrate, ambient temperature, 24 h Leachable
F2 : Acid-soluble Ammonium acetate, pH 5, ambient temperature, 5 h
Acid-soluble fraction, assumed carbonates
F3 : Reducible Hydroxylamine hydrochloride (3/4) in acetic acid (1/4), 96 °C, 6 h Mn and Fe oxyhydroxides
F4 : Oxidisable
Hydrogen peroxide and nitric acid, 96 °C, 2 h; then, hydrogen peroxide,
85 °C, ½ h.; and ammonium acetate in nitric acid addition at ambient
temperature
F5 : Residual Aqua regia Silicate-like quartz
At each step, concentrations are measured in the liquid extracted fraction by inductively coupled plasma atomic emission spectroscopy (ICP-AES) and mass spectroscopy, which provides the lowest detection limits. ICP-AES measurements are also applied on dried bulk sediment after chemical attack using aqua-regia. 3. Results 3.1. Chemical composition and mineralogical characterisation The sediment essentially consists of a silty (~ 40 %), clay (~ 20 - 25 %), carbonate (~ 20 %) and organic (~ 10 - 15 %) fractions. The XRD pattern presented in figure 1 indicates that sediments consist of major quartz and calcite with minor feldspar (albite, microcline), dolomite, and clay fraction dominated by illite, illite-smectite mixed layer, chlorite, and kaolinite, and iron sulphides like pyrite.
3
Figure 1. XRD pattern for the bulk sediment (surface sample). Mineralogical phases identified are quartz (Q), calcite (C), kaolinite (K), illite (I), chlorite (Ch), dolomite (D), hematite (H), pyrite (P), albite (A) and microcline (M).
3.2. Chemical sequential extractions Results are presented in figure 2 and table 2. Elemental concentrations of Zn, Pb and Fe measured in the bulk sediment are lower than in the sum of extracted fractions (figure 2). This point will be discussed in the following. Concentrations in the leachable fraction (table 1) are null (Pb) or too low (Fe, Zn) with regard to the others to be represented in the figure. Fe is mostly concentrated (~ 60 %) in the reducible fraction assumed to be Fe- and Mn-oxyhydroxide fraction. This fraction exhibits ~ 80 % of Pb and ~ 50 % of Zn. The acid-soluble (assumed carbonates) fraction concentrates ~ 10 % of Pb and 35 % of Zn. Table 2. Elemental concentrations (mg/kg) measured by ICP-AES in extracted fractions and bulk sediment and uncertainties (mg/kg). Detection Limits (DL) are reported for each element.
Extractants Zn (mg/kg) Pb (mg/kg) Fe (mg/kg) DL 1 5 2
Calcium nitrate 12 ± 1 < DL < DL Ammonium acetate 747 ± 20 62 ± 2 31 ± 4
Hydroxylamine hydrochloride 1331 ± 88 473 ± 11 17437 ± 544 Hydrogen peroxide, nitric acid 128 ± 17 58 ± 14 5760 ± 939
Aqua regia (residual) 36 ± 4 < DL 6394 ± 2154 Aqua regia (bulk sediment) 1685 ± 10 446 ± 14 17113 ± 251
4
Figure 2. Fe, Pb and Zn concentrations (mg/kg) measured by ICP-AES in extracted fractions and bulk sediment. Uncertainties are reported in table 1.
3.3. Micro-analyses by SEM-EDS, EPMA and µ-XRF Several zones of interest in embedded sample were observed by SEM. Backscattered electron images and EDS spectra are used for mineralogical phase’s identification. An example is presented in figure 3, resuming various phases encountered in the sediment.
Figure 3. BSE image and zooms (white squares) in the embedded sample.
5
Microanalyses and elemental mappings among zones of interest were carried out using EPMA (figure 4). Maps presented in the figure (HV = 15 kV, beam current = 15 nA, dwell time 100 ms/pixel, 256 x 256 pixels) give the distribution of Si, Al, Fe and Ca. The presence of micro-carbonates (calcite-like) distributed among the matrix, silicates (quartz), and an undetermined Al-, Ca-, Si-, Fe-rich particle are highlighted. No Zn or Pb signal was detected under these analytical conditions. The same region was analyzed by µ-XRF (see figure 5) with Zn and Pb detection limits (DL) below the EPMA ones. From these results, Zn and Pb seem to be diffusive among the matrix. No Ca-carbonate (calcite) correlations with Pb and Zn were underlined here.
Figure 4. Elemental distribution of Ca, Fe, Al and Si by EPMA (50 µm x 50 µm).
Quantitative EPMA analyses carried out on several calcite grains (45 points of analyses performed in 3 areas in a single thin section) showed Zn content in the range of 0 to 1478 mg/kg with a mean value of ~ 160 mg/kg. Furthermore, well crystallized calcites (inherited round grains, 20 - 100 µm size) are not correlated with detected Zn or Pb. It suggests that different carbonate generations (inherited well crystallized grains, authigenic micro-carbonates) coexist and only some type of calcium carbonate may carry some Zn or Pb in trace concentrations. Thus, in figure 3, a dolomite grain surrounded by a µ-Fe-oxyhydroxide corona can be observed. Dolomite does not present any Zn or Pb content, whereas the Fe-rich corona exhibit significant Zn and Pb amounts. Others Fe-rich phases carrying Zn and Pb have also been noticed: reduced phases such as Fe-sulphides (pyrite FeS2, figure 5). 4. Discussion The correlation between carbonates and trace elements (Zn and Pb), highlighted by chemical extraction, is not consistent with microanalyses on defined areas. Assuming 20 % of carbonates in the sample and a Zn concentration of 750 mg/kg measured in the acid-soluble extraction fraction (assumed to be the carbonate-rich fraction), the Zn amount attended in carbonates would be 3750 mg/kg. Yet, the detection of Zn in carbonates by EPMA is about 20 times lower (considering a
6
Figure 5. Ca, Fe, Zn, and Pb mappings (100 µm x 100 µm) by µ-XRF using synchrotron radiation.
Zn mean value in carbonates of 160 mg/kg). Some hypotheses can be done to explain this inconsistency: - Sequential chemical extraction may not be selective extractions: some phases, sensitive to
ammonium acetate attack, can enrich the acid-soluble fraction by dissolution (such as phosphate phases since 5 % of P is extracted in this fraction) and lead to Zn and/or Pb releasing.
- An unreasonable difference (up to 30 %) between bulk amounts of Fe, Zn and Pb and sums of concentrations in the 5 extraction fractions is observed. It can be explained by the use of aqua-regia to determine bulk concentration in a pseudo-total way, leading to an under-estimation of global concentrations in samples. Moreover, samples heterogeneity (in chemical compositions, spatial distribution on the field, crystalline structures, grain sizes, nugget effects for example) can induce different types of contacts between reactants and reagents involving various dissolution pathways. Another chemical extraction procedure is in progress, in order to determine compositions by
ICP-AES and µ-XRF and to compare them respectively in liquid extracted fractions and solid residues of each extraction step. It will allow us determining whether the carbonate fraction is totally attacked and solubilized or whether the recalcitrant forms of carbonates remain in other fraction(s) (ankerite Ca(Fe2+,Mg,Mn2+)(CO3)2, siderite FeCO3, …). Another scope will be to estimate the potential selectivity of extractants. For instance, we’ll study if some phases are attacked concomitantly with carbonates during the acid-soluble attack. In the same way, the use of aqua-regia and alkaline fusion protocols to extract total concentrations and validate the pseudo-total behaviour of aqua-regia extractant will be tested.
The correlation of Fe with Zn and/or Pb has been demonstrated among mineralogical phases under different oxidation states. Fe speciation may be a key parameter on the dynamic of trace elements mobility. As demonstrated in this study, Fe-oxyhydroxides and Fe-sulphides are Zn and Pb carriers.
The fine grained matrix, assumed to be silicates and clays but also amorphous phases like organic matter, can be trace element carriers by adsorption or interfoliar substitutions on clays or creation of organo-minerals complexes.
7
5. Conclusion The characterisation of the global mineralogy of the sediment and identification of trace metal carriers has been carried out by combining different approaches: chemical extraction analysis and spectroscopic techniques such as SEM-EDS, EPMA, XRD, XRF and synchrotron radiation µ-XRF.
Chemical sequential extractions carried out on the bulk sediment show a high Zn content in the reducible extracted fraction and a minor but significant Zn content in the acid-soluble extracted fraction (~ 3750 mg/kg) assumed to be carbonates. Pb is more concentrated in the reducible fraction than in the acid-soluble fraction. Microanalysis using EPMA and synchrotron radiation µ-XRF show a lower Zn content in carbonates (about 20 times lower), and a well correlated Zn and Pb association with Fe and a fine grained silicate matrix enriched in Zn and Pb.
On the one hand, the trapping capacity of the carbonate fraction, expected by the chemical extraction, is not validated by microanalysis carried out on inherited and well crystallized carbonates, nor on neo-formed and micro-sized ones. This discrepancy could be explained by dissolution of unexpected mineralogical phases during the acid-soluble extraction step leading to reassign Zn and Pb initial distribution to phases soluble at ammonium acetate, like phosphates. Nevertheless, the study of solid residues, compared with results on liquid extracted fractions, is in progress to establish a geochemical distribution balance. On the other hand, Zn-Fe and Pb-Fe associations had been revealed among phases like Fe-oxyhydroxides and Fe-sulphides. Finally, data processing of Fe speciation will be pointed out by XAS analysis carried out on Fe and Zn-rich phases, especially by studying XANES spectra at Fe Kα edge. Acknowledgments This work was financially supported by the BRGM and the Regional Council of the Région Centre. The authors are also grateful to the MMA (BRGM) and I18 beamline (Diamond Light Source) staffs for technical help and to E. Veron (CEMHTI) for the SEM observations. References [1] Tessier A, Campbell P G C and Bisson M 1979 Anal. Chem. 51 844-51 [2] Gomez-Ariza J L, Giràldez I, Sànchez-Rodas D and Morales E 1999 Int. J. Environ. Anal.
Chem. 414 151-64 [3] Leleyter L and Baraud F 2005 C. R. Geoscience 337 571-79 [4] Isaure M P, Laboudigue A, Manceau A, Sarret G, Tiffreau C, Trocellier P, Lamble G,
Hazemann J L and Chateigner D 2002 Geochim. Cosmochim. Acta 66 1549-67 [5] Kirpichtchikova T A, Manceau A, Spadini L, Panfili F, Marcus M A and Jacquet T 2006
Geochim. Cosmochim. Acta 70 2163-90 [6] Sarret G, Balesdent J, Bouziri L, Garnier J M, Marcus M A, Geoffroy N, Panfili F and
Manceau A 2004 Environ. Sci. Technol. 38 2792-801 [7] Manceau A, Lanson B, Schlegel M L, Harge J C, Musso M, Eybert-Bérard L, Hazemann J L,
Chateigner D and Lamble G M 2000 Am. J. Sci. 300 289-343 [8] Pouchou J L and Pichoir F 1991 Quantitative analysis of homogeneous or stratified
microvolumes applying the model “PAP”. in: Electron probe quantitation. Heinrich K F J and Newbury D E, eds. (New York: Plenum Press) 31-75
[9] Ancey M, Bastenaire F and Tixier R 1978 Application des méthodes statistiques en microanalyse. Ecole d’été de Microanalyse et de Microscopie Electronique à Balayage. (St. Martin d’Heres: Editions de Physique)
[10] Piou S, Bataillard P, Laboudigue A, Férard J F and Masfaraud J F 2009 Environ. Res. 109 712-20
8

Combining trace elements micro-analysis in deposited dredged

Documents