ORIGINAL PAPER Determination of mercury species ...

3 downloads 0 Views 353KB Size Report
ception of cinnabar crystals), but are chemically or physically bound to the surface of particles. Thus, the binding interactions of analytes in the matrix and.
Chemical Papers 68 (4) 427–434 (2014) DOI: 10.2478/s11696-013-0471-0

ORIGINAL PAPER

Determination of mercury species using thermal desorption analysis in AAS Pavel Coufalík, Ondřej Zvěřina, Josef Komárek* Department of Chemistry, Faculty of Science, Masaryk University, Kotlářská 2, Brno 611 37, Czech Republic Received 1 May 2013; Revised 18 June 2013; Accepted 19 June 2013

Analytical aspects of the determination of inorganic mercury (Hg) species by thermal desorption followed by atomic absorption spectrometry (AAS) detection were investigated in this work. Characteristic Hg release curves of the following species were observed: Hg0 , HgCl2 , HgO, HgSO4 , HgS, and the Hg bound to humic acids. Particular attention was dedicated to the thermal stability and change of bond of Hg0 in the following matrices: sand, kaolinite, granite, peat, power plant ash, and soil. The bond of elemental Hg in environmental materials was described on basis of this experiment. Contaminated soil samples from two locations in the Czech Republic were investigated by thermal desorption analysis. Afterwards, the contents of volatile and plant-available Hg in the studied samples were determined. The determination of Hg0 using the thermal method was related to the results of liquid sequential extraction. The development of Hg speciation and the stability of Hg were assessed on basis of the data obtained. Thus, the analytical procedure used is a suitable tool for the study of inorganic Hg species in contaminated soils. c 2013 Institute of Chemistry, Slovak Academy of Sciences  Keywords: mercury, speciation, soil, thermal desorption, extraction

Introduction The amount of Hg circulating in ecosystems represents a serious problem nowadays. Contamination of anthropogenic origin substantially increases the quantity of Hg which is transported between atmosphere and surface, as well as the total amount of this metal and its compounds deposited in oceans, sediments and soils. Toxicity of mobile Hg forms is a distinct biospherical threat, especially in case of organometallic species formed via biological processes from inorganic Hg forms. The process of speciation change is dependent on both environmental conditions and properties of the original form. The study of Hg speciation in soil and similar materials constitutes a form of trace analysis, which is essential for the risk assessment of contaminated environmental samples. Natural Hg species almost never occur freely scattered in the material (with the exception of cinnabar crystals), but are chemically or

physically bound to the surface of particles. Thus, the binding interactions of analytes in the matrix and their stability should be the main objects of study. Besides sequential extractions (Issaro et al., 2009), thermal desorption analysis from the solid phase has been promoted (Biester & Nehrke, 1997; Bollen & Biester, 2011). This method provides information about the binding intensity of species in the matrix and about the contents of certain Hg forms such as Hg0 and HgS. Determination of these species by means of thermal desorption analysis may be even more successful (Biester & Scholz, 1996) than that by liquid extractions. Nevertheless, thermal desorption analysis is not used by many authors and its interpretation is still a question of debate. In this work, the primary subject is thermal desorption analysis and its application in the speciation analysis of inorganic Hg forms. The method is based on the low thermal stability of Hg compounds, where the studied samples are heated in a stream of inert

*Corresponding author, e-mail: [email protected]

Unauthenticated Download Date | 11/24/15 8:01 AM

428

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

gas (Ar, N2 ) from ambient temperature up to 800 ◦C in some cases (Feng et al., 2004). The released Hg is measured on-line in an AAS optical system (the most commonly used method) during sample heating. Thereafter, the recorded signal is converted to a graph of absorbance versus temperature, the so-called mercury release curve. The main intention of this research was to determine characteristic desorption curves for the most common Hg forms. Furthermore, the application of thermal desorption analysis to the study of contaminated soil samples was solved. The second aim was to determine the levels of elemental Hg using thermal desorption and to confront the results with data obtained by means of an extraction procedure. Thus, soil samples were subjected to liquid extraction to assess the conformity of the results of both methods and to determinate the plant-availability of the present Hg.

Experimental Contaminated soil samples were collected at two locations in the Czech Republic. The first location (samples F3A–F5E) was in the vicinity of an incinerator in Hradec Králové (50◦ 13 16 N, 15◦ 49 4 E), which was closed in 2002. The second site (samples 1–16) was in the proximity of a former mill in the village of Trhové Dušníky (49◦ 43 7 N, 14◦ 0 54 E). The staining of corn by phenyl mercury chloride was carried out in this location. The collection of samples was conducted using a stainless sampling rod to the depth of 25 cm. The samples were dried at ambient temperature and sieved to below 2 mm. The collected material was stored at 4 ◦C in PE flasks until analysis. A portion of each sample was ground in a Pulverisette 7 mill to < 63 m. Artificial calibration materials with HgCl2 , HgO, HgSO4 , and HgS (Sigma–Aldrich, Germany) were prepared by the dilution of solid compounds with sea sand (Penta, Czech Republic); the material with Hg0 was prepared by adsorption of Hg0 droplets on sand particles. Hg bound to humic substances was prepared by soaking an acid solution (pH = 3) of HgCl2 through pure humic acids (Sigma–Aldrich, Germany) in a column for 72 h (Coufalík et al., 2012). The humic acids were dried at ambient temperature and homogenised in a mortar. Then, the final Hg content in humic acids was determined. In addition, materials enriched with Hg0 were prepared. Sea sand, ground granite, kaolinite, peat, power plant ash, and one soil sample (sample 16, Trhové Dušníky) were incubated with Hg0 vapours in an argon atmosphere in a desiccator for three months. The materials were scattered on Petri dishes in thin layers. A droplet of Hg (1 g) on a watch glass was placed at the bottom of a desiccator. The materials were used without any modification; these samples were dry, with the particle size below 2 mm. The percent-

age of particles below 63 m was determined for these materials by sieving. Hg contents in the samples were measured before and after the incubation experiment. Thermal desorption analysis of Hg in the artificial materials and soil samples was carried out using a prototype of a heating unit connected with a Perkin– Elmer 3030 AA spectrometer with deuterium background correction (Coufalík et al., 2012). Argon was used as a carrier gas and warming medium. Mercury release curves were observed from 50 ◦C up to 450 ◦C. The temperature gradient was 30 ◦C min−1 . The released Hg was measured on-line in a quartz tube which was heated to 800 ◦C to ensure complete atomisation (e.g. the desorption intervals of HgO and HgSO4 are below the temperature of decomposition, and atomisation at 800 ◦C is essential during the Hg determination by AAS). The flow through the cuvette was 500 mL min−1 . The system was designed for a dosage of 500 mg of the sample with steady heating of the material according to the temperature gradient. The sample was placed in a cylindrical sintered glass capsule together with layers of quartz wool (Merck, Germany). Hg release curves for the soil samples were measured for the same Hg contents in absolute values (100 ng of Hg). Thus, the different mass of samples allowed an accurate comparison of the Hg thermal stability despite significant variations in the overall contents. All soil samples were also heated in an oven with a ventilation system (Memmert UFE 400) at 105 ◦C for 48 h. The difference between the total Hg content and the content after thermal treatment can be considered as released Hg0 (Nóvoa-Mu˜ noz et al., 2008). It is obvious that the thermal release of highly volatile organic Hg forms also takes place under these conditions. Sequential extraction (Coufalík et al., 2012) was also used for the analysis of Hg in soil samples. The first fraction of this procedure is water-soluble Hg as the most mobile form (the second and third fractions are defined as the Hg releasable under acidic conditions and humic bound Hg, respectively). Extraction with 50 % HNO3 (the fourth extraction step) yields Hg0 and Hg bound in complexes and amalgams. Thus, the content of Hg in this fraction should be equal to or higher than the thermally desorbed Hg0 at 105 ◦C. In addition, extraction with a solution of 1 M NH4 NO3 (Merck, Germany) according to standard DIN 19730 (Deutsches Institut f¨ ur Normung, 1997) was also performed for the estimation of the plant-available Hg portion. The measurement of total Hg contents in solid samples as well as in liquid extracts was performed on an AMA-254 analyser (Altec, Czech Republic). This is an atomic absorption spectrometer employing thermooxidative decomposition of the sample and a preconcentration of Hg on the amalgamator (analogous to the DMA-80 analyser). The total Hg contents in Hg0 incubated materials were determined from the areas

Unauthenticated Download Date | 11/24/15 8:01 AM

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

429

Fig. 1. Hg release curves for pure Hg species: Hg0 (1), HgCl2 (2), HgO (3), HgSO4 (4), humic bound Hg (5), HgS (6).

of peaks using thermal desorption analysis.

Results and discussion Verification of measurement parameters of individual Hg compounds Instrumentation used for the thermal desorption analysis allowed precise characterisation of the Hg release curves for pure Hg species. An appropriate range of measurable concentrations in relation to the broadening of the peak with the increasing absorbance (A) was determined for each Hg form. Limits of quantification were calculated for each form according to the peak area at the peak height of A = 0.004 for the highest dosage of the sample (500 mg). The amount of Hg at the peak height of 0.004 represents the lowest Hg content which could be determined with analytical accuracy. Measurements of desorption intervals were performed at the peak height of 0.012 (not at higher absorbances) in order to achieve correct resolution of individual species. Broadening of the peak is considerable at high absorbance values and the characteristic intervals cannot be determined. The same peak height was reached for different mass of the sample. Identical height of the peaks is necessary for the comparison of the Hg forms observed. Fig. 1 represents the thermal curves of all observed Hg forms with the following desorption intervals: 60–120 ◦C for Hg0 (maximum at 95 ◦C), 100–190 ◦C for HgCl2 (maximum at 140 ◦C), 110–400 ◦C for HgO (maximum at 240 ◦C), 165–450 ◦C for HgSO4 (maximum at 255 ◦C), 270–340 ◦C for Hg bound to humic acids (maximum at 305 ◦C) and 250– 435 ◦C for HgS (maximum at 365 ◦C). The shape and broadening of the peaks of Hg forms are dependent on the concentration of the species. Shapes of the desorption curves up to A = 0.5 were studied during the calibration. Characteristic mercury release curves are significantly influenced by the

preparation of artificial materials. However, tailing, fronting, or other irregularities of the peak are typical of each form at higher concentrations (Fig. 2). The limit of quantification of the species is then derived from the range of the intervals of the desorption temperatures. Relative standard deviation (RSD) of the peak areas was up to 14 % for all concentrations of the observed species (three parallel analyses). The Hg release curves of individual species in Fig. 2 are in the following range of g Hg: Hg0 – 0.044–0.684 g Hg; HgCl2 – 0.098–3.23 g Hg; HgO – 0.153–2.08 g Hg; HgSO4 – 0.100–1.34 g Hg; humic bound Hg – 0.074–3.15 g Hg; HgS – 0.113–3.73 g Hg. Low thermal stability of Hg0 adsorbed on sand particles caused a significant drift in the signal at concentrations higher than 1 mg kg−1 in some cases. In extreme cases, splitting of the peak can be observed for weakly adsorbed Hg on particle surfaces and for Hg in pores. In this case, homogenisation of the materials prior to the measurement is essential. Rearrangement of Hg atoms in the material was observed during the homogenisation due to the volatility of Hg0 under standard conditions. An example of a well-prepared material is presented in Fig. 2a. Elemental Hg is susceptible to oxidation in the presence of a halogen compound (Windm¨ oller et al., 1996). Generally, a double-peak can be caused by poor preparation of the material or by unexpected speciation changes, which can be however diminished by sample homogenisation. The peak of Hg0 in Fig. 1 is narrow and well resolved because of the low Hg concentration. It is the desorption curve of 36 ng of Hg. The limit of quantification for this form was 0.024 mg kg−1 and it was the lowest of all studied species due to the narrow range of the desorption interval. Mercury chloride (HgCl2 ) showed a shift in the signal to higher desorption temperatures in the presence of particles with strong adsorptive capability such as clay (Coufalík et al., 2012). In case of the weak ad-

Unauthenticated Download Date | 11/24/15 8:01 AM

430

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

Fig. 2. Thermal behaviour of Hg0 (a), HgCl2 (b), HgO (c), HgSO4 (d), humic bound Hg (e), and HgS (f).

Unauthenticated Download Date | 11/24/15 8:01 AM

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

431

Fig. 3. Hg release curves of incubated materials: sand (1), kaolinite (2), ash (3), soil (4), peat (5), granite (6).

sorptive capability of sand, the peak fronting and tailing occurred only at higher concentrations (Fig. 2b). The limit of quantification (LOQ) for HgCl2 was 0.035 mg kg−1 . Mercury oxide should only be measured at low absorbances because of signal dispersion over the wide range of desorption temperatures (Fig. 2c). LOQ for HgO was 0.1 mg kg−1 . Desorption intervals of HgO and HgSO4 occur at higher temperatures, according to some studies (Biester et al., 1999, 2000; Hojdová et al., 2008, 2009). Nevertheless, the published results were related to high concentrations of Hg up to the absorbance of 1.8. Thermal desorption of HgO and HgSO4 at the measured temperature intervals (Fig. 1) was mainly caused by low concentrations of Hg (151 ng and 99 ng of Hg) and by the perfect homogenisation of the materials with microcrystalline forms of substances. LOQ for HgSO4 was 0.066 mg kg−1 . The most symmetrical signal across all absorbances (Fig. 2e) proved the Hg bound to humic substances with an LOQ of 0.026 mg kg−1 . Peak fronting appeared during the thermal desorption of cinnabar with the increasing Hg concentrations (Fig. 2f). Nonetheless, a steep signal increase occurred after the breakdown of the crystal lattice. LOQ for HgS was 0.061 mg kg−1 . It is appropriate to implement the determination of individual Hg forms at the lowest concentrations of analytes to achieve the best resolution (peak overlapping is obvious due to the wide desorption intervals of the species and the matrix of the studied materials). Thermal stability of incubated materials Adsorption of Hg vapour and its thermal stability in the different matrices was studied. Selection of materials was carried out according to the following criteria: size and porosity of the adsorption surface (sand, kaolinite), different content of organic matter (peat, granite), and artificial or natural origin of complex matrices (power plant ash, soil sample). The performed experiment enabled characterisation of the sta-

bility of free elemental Hg in the matrix under controlled conditions. Hg release curves in the investigated matrices are presented in Fig. 3. RSD of the peak areas was up to 9 % for three parallel analyses. The RSD achieved is relatively low in comparison with that presented in literature (Biester & Scholz, 1996) despite the volatility of the adsorbed Hg0 . Introduced Hg0 usually has a higher RSD than elemental Hg, which is created by the secondary reduction of Hg2+ (Biester & Scholz, 1996). High homogeneity of the Hg contents was probably caused by a long incubation period related to the perfect saturation of the binding sites. Incubated Hg0 on sand (Fig. 3, curve 1) proved higher thermal stability (peak at 125 ◦C) compared with a simple shaking of Hg0 in the sand. The thermal shift occurring during the incubation cannot be attributed to oxidation due to the inert atmosphere. Therefore, the shift is probably caused by perfect capture of Hg at binding sites; the measured interval and the temperature of maximum release (125 ◦C) correspond to data in literature (Biester & Zimmer, 1998). The total Hg content in the sand was only 0.054 mg kg−1 due to the limited possibilities of sorption (no particles below 63 m). The desorption maximum of Hg0 in kaolinite (Fig. 3, curve 2) was set to 158 ◦C owing to the adsorption surface. The atoms of Hg0 can be trapped in very fine mineral particles (below 63 m) and thus increase its thermal stability. The total Hg content in kaolinite was 57 mg kg−1 . The second parameter observed was the influence of organic matter. Hg trapped in ground granite (curve 6) proved the highest thermal stability of the studied materials. The desorption interval of 200–450 ◦C indicates the oxidation of Hg0 . Hg captured in peat (curve 5) showed a temperature range of 150–400 ◦C. The shape and position of the peak are indicative of a bivalent Hg form (do Valle et al., 2006). However, the binding character of Hg in this material (physical adsorption x chemisorption) could not be determined in this case. Generally, Hg can also be bound by adsorp-

Unauthenticated Download Date | 11/24/15 8:01 AM

432

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

Table 1. Contents of total and volatile Hg and Hg fractionation Content of Hg/(mg kg−1 ) Sample Totala F3A F3B F3C F3D F3E F5A F5B F5C F5D F5E 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1.07 0.236 0.415 0.419 0.55 2.05 0.396 0.58 28.8 10.5 2.11 0.85 1.93 1.34 1.33 1.08 1.43 1.14 1.04 1.56 0.89 3.86 0.97 2.5 9.8 1.10

Volatileb

Extractablec

Extractabled

Plant-availablee

0.32 0.098 0.091 0.100 0.27 0.82 0.094 0.077 12.2 4.0 0.43 0.143 0.71 0.36 0.42 0.34 0.45 0.22 0.25 0.30 0.23 0.59 0.159 0.44 1.29 0.169

0.61 0.13 0.25 0.24 0.32 1.4 0.28 0.35 16.2 4.9 0.80 0.023 0.76 0.16 0.28 0.25 0.47 0.17 0.11 0.49 0.21 1.4 0.13 0.51 4.9 0.13

– – – – – – – – 0.40 0.156 0.023 – 0.0075 0.0174 – 0.0093 – – – – – 0.0156 – 0.0173 0.026 –

– – – 0.0011 – – – – 0.052 0.025 0.0012 0.0014 0.0009 – – – – 0.0007 – – – 0.0021 – 0.0019 0.0017 –

a) RSD < 2 %; b) RSD < 5 %; c) extractable with 50 % HNO3 ; RSD < 10 %; d) extractable with water; RSD < 5 %; LOQ = 0.002 mg kg−1 ; e) RSD < 8 %; LOQ = 0.0007 mg kg−1 .

tion (as on mineral surfaces) after the saturation of the binding sites on humic acids (Biester & Zimmer, 1998). Although both materials had the same content of particles below 63 m (7.5 %), the Hg content in peat was 42 mg kg−1 while that in granite was 48 mg kg−1 . Therefore, the mineral phase seems to be a better adsorbent in this experiment. Progress of the oxidation of Hg0 onto mineral particles was recorded on the Hg release curve of power plant ash (curve 3). The first peak at 150 ◦C corresponds to the elemental form integrated in the matrix; the second peak at 300 ◦C represents already oxidised Hg. Hg content in the ash was 65 mg kg−1 . The highest Hg enrichment was observed for the soil sample (curve 4). The increase of total Hg content during the incubation was in the range of 1.1–236 mg kg−1 . Thus, a complex matrix of soil with organic and mineral components in the natural state has the highest sorption capacity despite the low content of particles below 63 m (6 % in soil and in ash). The oxidation of Hg0 took place and no free elemental Hg was recorded after 90 days. Analysis of contaminated soil samples Total Hg contents were determined in the samples of contaminated soil with an LOQ of 0.3 g kg−1 . The

contents are shown in Table 1. RSD for these values was within 2 % (three parallel analyses). The samples were characterised by a significant range of concentration values within each location. The range of contents for the samples from the vicinity of the incinerator in Hradec Králové was 0.236–28.8 mg kg−1 . This is typical of point source contamination such as a waste dump which has a toxic impact on the environment (Mukherjee et al., 2004). Hg contents of the samples from Trhové Dušníky were in the range of 0.85–9.8 mg kg−1 . For the measurements of the Hg release curves of all samples, the mass corresponding to 100 ng of Hg in the absolute value was used due to significant differences in the total Hg contents. Fig. 4 shows the thermal desorption curves of samples from both locations. Desorption intervals were the same for all samples within each location. The peak in the range of 150–350 ◦C is typical of the samples from Hradec Králové (curves 1 and 2); the samples from Trhové Dušníky showed a desorption interval in the range of 170–360 ◦C (curve 1 is related to sample F5D; due to the necessity of sufficient mass of contaminated sample, the desorption of 300 ng of Hg in absolute value was carried out). Hg release curves of all the samples proved a single desorption peak which can be defined as a matrix nonspecific signal. Thus, free volatile Hg seemes not to be present in any samples.

Unauthenticated Download Date | 11/24/15 8:01 AM

433

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

a different method. Approximately half of the samples from this area also contained water-soluble and plantavailable Hg.

Conclusions

Fig. 4. Hg release curves of soil samples: Hradec Králové (sample F5D – curve 1; sample F3B – curve 2); Trhové Dušníky (sample 10 – curve 3).

This simple peak is probably caused by the strong influence of the matrix and by the amorphous nature of Hg species in the environmental sample. The absence of a characteristic peak below 150 ◦C during the thermal analysis does not mean the absence of Hg0 which is strongly bound in the matrix. The volatile fraction of contained Hg was determined from the difference in the contents before and after the heating at 105 ◦C for 48 h (Table 1). RSD was up to 5 % for three parallel analyses. Volatile Hg contents for the samples from Hradec Králové (samples F3A–F5E) ranged from 13 % to 49 %. The volatile fraction is defined as Hg0 plus organic forms, and the values were always lower than the amount of Hg in the extract of 50 % HNO3 . Hg contents of organometallic species are usually low in most soils (Schl¨ uter, 2000) with the exception of source contamination by these forms. The determination of Hg0 after heating at 105 ◦C can be considered as relevant in this case since a higher content of thermally determinated Hg0 can be caused by the desorption of volatile organic species or by a reduction process occurring during sample heating. The amounts of water-soluble (highly mobile) and plantavailable Hg were below the limit of quantification (2 g kg−1 and 0.7 g kg−1 , respectively) for most of the samples from this location. Volatile fraction of Hg in the samples from Trhové Dušníky (samples 1–16) was 13–37 %. However, for seven samples, this proportion was higher than the amount of Hg extracted in 50 % HNO3 during sequential extraction. Especially high difference was observed for sample 2 (Table 1). The volatile fraction was determined to be 16.8 % while the liquid extract formed only 2.7 % of the total Hg content. It is evident that this disagreement results from the presence of organic Hg species since residues can be expected in this area. However, their determination should be carried out by

Hg contained in environmental samples is subjected to speciation changes based on external conditions. These changes can be studied by means of thermal desorption analysis from the entrance of contamination to the stable binding of different Hg forms in the material. Simulation of speciation changes during the incubation experiment with Hg0 revealed important information about the elemental Hg determination in the material. The absence of a peak at around 100 ◦C on the Hg release curve does not invalidate the presence of Hg0 in the material. This follows from a comparison of the results of thermal desorption analysis, simple desorption at 105 ◦C, and liquid extraction. The effect of Hg stabilisation by oxidation, adsorption on internal surfaces, or through speciation changes is evident in the studied contaminated samples. Although a matrix nonspecific signal does not allow direct evaluation of the content of Hg0 in this case, it indicates advanced speciation in the contaminated material and causes low contents of mobile Hg forms (water-soluble and plant-available fractions). Although the Hg contained is integrated in the soil matrix, potential remobilisation into the environment is not prevented. Thus, the degree of contamination of the soil samples was assessed. The soils were relatively stable, without any free Hg0 and with low amounts of mobile and plant-available Hg. Nevertheless, the locations from which these samples originate can be considered as hazardous to the environment due to the high content of total Hg and potential occurrence of organometallic compounds. Acknowledgements. The authors are grateful for financial support from the Grant Agency of the Czech Republic, project P503/12/0682 and the Masaryk University in Brno, project MUNI/A/0969/2012.

References Biester, H., & Scholz, C. (1996). Determination of mercury binding forms in contaminated soils: Mercury pyrolysis versus sequential extractions. Environmental Science & Technology, 31, 233–239. DOI: 10.1021/es960369h. Biester, H., & Nehrke, G. (1997). Quantification of mercury in soil and sediments – acid digestion versus pyrolysis. Fresenius’ Journal of Analytical Chemistry, 358, 446–452. DOI: 10.1007/s002160050444. Biester, H., & Zimmer, H. (1998). Solubility and changes of mercury binding forms in contaminated soils after immobilization treatment. Environmental Science & Technology, 32, 2755–2762. DOI: 10.1021/es9709379. Biester, H., Gosar, M., & M¨ uller, G. (1999). Mercury speciation in tailings of the Idrija mercury mine. Journal of Geochemical Exploration, 65, 195–204. DOI: 10.1016/s0375-

Unauthenticated Download Date | 11/24/15 8:01 AM

434

P. Coufalík et al./Chemical Papers 68 (4) 427–434 (2014)

6742(99)00027-8. Biester, H., Gosar, M., & Covelli, S. (2000). Mercury speciation in sediments affected by dumped mining residues in the drainage area of the Idrija mercury mine, Slovenia. Environmental Science & Technology, 34, 3330–3336. DOI: 10.1021/es991334v. Bollen, A., & Biester, H. (2011). Mercury extraction from contaminated soils by L-cysteine: Species dependency and transformation processes. Water, Air, and Soil Pollution, 219, 175–189. DOI: 10.1007/s11270-010-0696-2. Coufalík, P., Krásenský, P., Dosbaba, M., & Komárek, J. (2012). Sequential extraction and thermal desorption of mercury from contaminated soil and tailings from Mongolia. Central European Journal of Chemistry, 10, 1565–1573. DOI: 10.2478/s11532-012-0074-6. Deutsches Institut f¨ ur Normung (1997). German standard: Bodenbeschaffenheit – Extraktion von Spurenelementen mit Ammoniumnitratl¨ osung. DIN 19730. Berlin, Germany. (in German) do Valle, C. M., Santana, G. P., & Windm¨ oller, C. C. (2006). Mercury conversion processes in Amazon soils evaluated by thermodesorption analysis. Chemosphere, 65, 1966–1975. DOI: 10.1016/j.chemosphere.2006.07.001. Feng, X., Lu, J. Y., Gr`egoire, D. C., Hao, Y., Banic, C. M., & Schroeder, W. H. (2004). Analysis of inorganic mercury species associated with airborne particulate matter/aerosols: method development. Analytical and Bioanalytical Chemistry, 380, 683–689. DOI: 10.1007/s00216-004-2803-y. Hojdová, M., Navrátil, T., & Rohovec, J. (2008). Distribution and speciation of mercury in mine waste dumps. Bulletin of Environmental Contamination and Toxicology, 80, 237–241. DOI: 10.1007/s00128-007-9352-y.

Hojdová, M., Navrátil, T., Rohovec, J., Penížek, V., & Grygar, T. (2009). Mercury distribution and specition in soils affected by historic mercury mining. Water, Air, and Soil Pollution, 200, 89–99. DOI: 10.1007/s11270-008-9895-5. Issaro, N., Abi-Ghanem, C., & Bermond, A. (2009). Fractionation studies of mercury in soils and sediments: A review of the chemical reagents used for mercury extraction. Analytica Chimica Acta, 631, 1–12. DOI: 10.1016/j.aca.2008.10.020. Mukherjee, A. B., Zevenhoven, R., Brodersen, J., Hylander, L. D., & Bhattacharya, P. (2004). Mercury in waste in the European Union: sources, disposal methods and risks. Resources, Conservation and Recycling, 42, 155–182. DOI: 10.1016/j.resconrec.2004.02.009. Nóvoa-Mu˜ noz, J. C., Pontevedra-Pombal, X., Martínez-Cortizas, A., & García-Rodeja Gayoso, E. (2008). Mercury accumulation in upland acid forest ecosystems nearby a coalfired power-plant in Southwest Europe (Galicia, NW Spain). Science of the Total Environment, 394, 303–312. DOI: 10.1016/j.scitotenv.2008.01.044. Schl¨ uter, K. (2000). Review: evaporation of mercury from soils. An integration and synthesis of current knowledge. Environmental Geology, 39, 249–271. DOI: 10.1007/s002540050005. Windm¨ oller, C. C., Wilken, R. D., & De Figueiredo Jardim, W. (1996). Mercury speciation in contaminated soils by thermal release analysis. Water Air and Soil Pollution, 89, 399–416. DOI: 10.1007/bf00171644.

Unauthenticated Download Date | 11/24/15 8:01 AM