Lead isotope ratio measurements by ICP-QMS to identify ... - CiteSeerX

Science of the Total Environment 367 (2006) 988 – 998 www.elsevier.com/locate/scitotenv

Lead isotope ratio measurements by ICP-QMS to identify metal accumulation in vegetation specimens growing in mining environments E. Marguí a , M. Iglesias a , I. Queralt b , M. Hidalgo a,⁎ b

a Department of Chemistry, University of Girona, Campus Montilivi, 17071 Girona, Spain Institute of Earth Sciences “Jaume Almera”, CSIC, Solé Sabarís s/n, 08028 Barcelona, Spain

Received 4 October 2005; received in revised form 17 March 2006; accepted 17 March 2006 Available online 15 May 2006

Abstract The use of variations in stable Pb isotope ratios has become a well-established diagnostic technique for characterising sources of lead contamination. In this work, lead isotope ratios in mining wastes (lead content 320–130,000 mg kg− 1) and vegetation specimens (lead concentration 7–650 mg kg− 1) have been determined by inductively coupled plasma quadrupole-based mass spectrometry (ICPQMS) in order to investigate lead bioaccumulation in Buddleia davidii growing on wastes from two abandoned Pb/Zn mining areas in Spain. The accuracy of the isotope ratio measurements was evaluated by analysing a certified isotopic standard NIST SRM 981. Good agreements were obtained between the lead isotope ratios measured and the certified values (deviations within 0.01–0.2%). The results indicate that the lead isotopic ratios in vegetation samples collected in the mining areas differed from those of a specimen from an uncontaminated site (control sample). However, close lead isotope ratio values were found between vegetation specimens and mining tailings. Therefore, the results suggest that lead in the collected vegetation specimens is most likely related to the influence of mining activities rather than to other sources like past leaded-petrol emissions. © 2006 Elsevier B.V. All rights reserved. Keywords: Lead isotope ratios; Mining activity; Buddleia davidii; ICP-QMS

1. Introduction Lead pollution of the environment as a result of human activities is an acute problem nowadays. Its high persistence and its tendency to accumulate in living organisms have promoted lead as the target of most current environmental studies (Elbering et al., 2002; Prego and Cobelo-García, 2004; Vinagre et al., 2004). ⁎ Corresponding author. Tel.: +34 972418190; fax: +34 972418150. E-mail address: [email protected] (M. Hidalgo). 0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2006.03.036

However, the wide range of activities involving the release of this metal often makes lead source identification very difficult, but necessary to plan effective abatement strategies in contaminated areas (Spiro et al., 2004). The isotopic composition of lead is variable in nature as only one (204 Pb) of the four stable isotopes of this element (204Pb, 206Pb, 207 Pb and 208 Pb) is not the end product of the decay of thorium and uranium isotopes. So, although primordial lead had a fixed isotopic composition, the amounts of 206Pb, 207 Pb and 208Pb have

E. Marguí et al. / Science of the Total Environment 367 (2006) 988–998

increased over time relative to 204Pb (Platzner, 1997). Taking this into account, the use of variations in lead stable isotope ratios has become a well-established diagnostic technique to characterise sources of Pb contamination and has been applied to evaluate the global impact of various anthropogenic sources in different environmental matrices including sediments (Négrel and Roy, 2002; Négrel et al., 2004), soils (Bacon and Dinev, 2005), river water (Kurkjian et al., 2004), tap water (Cheng and Foland, 2005), peat bogs (Jackson et al., 2004), tree rings (Bellis et al., 2004), plants (Rabinowitz, 1972) and mussels (Labonne et al., 2001). Isotope ratio measurements have traditionally been carried out using thermal ionisation mass spectrometry (TIMS). This technique provides the precision and accuracy required in isotopic analysis. Nevertheless, exhaustive sample preparations, tedious sample handling protocols, long analysis periods (2 to 24 h), and inflexible analytical routines have made the use of TIMS unfeasible in some environmental studies where results are drawn from a large number of samples. For this reason, inductively coupled plasma mass spectrometry (ICP-MS) has been increasingly used in isotope ratio measurements (Al-Ammar and Barnes, 2001; Ettler et al., 2004; Xie and Kerrich, 2002; Boulyga et al., 2001) in recent years. Although ICPMS analysis cannot achieve the precision and accuracy in isotope ratio determination attained in TIMS, several advantageous features make the use of ICP-MS a suitable alternative: fast sample throughput, low sample analysis cost, instrument robustness and simplified sample preparation.

989

This work highlights the suitability of stable lead isotope ratios determined by quadrupole-based ICP-MS (ICP-QMS) to study the potential sources of the anomalous lead content in vegetation specimens (Buddleia davidii) growing on the wastes and soils of two abandoned Pb/Zn mining areas in Spain. Circumstantial evidence pointed strongly to contamination from mining activities as the prime source of lead but sources such as past leaded-petrol emissions could not be excluded. Therefore, lead isotope ratios in B. davidii were compared to those obtained in the isotopic analysis of the different kinds of wastes collected at the areas of study. The determined isotopic composition was also compared to that measured in several natural soils and in a leaded-petrol sample, in order to assess the most probable source of the accumulation of lead in vegetation specimens. 2. Experimental 2.1. Study areas and sampling Two abandoned Pb–Zn mining areas in Spain (Fig. 1) were studied in this work. The first is located in the north Pyrenean Range (Val d'Aran) whereas the second mining district is situated in the “La Selva” region (NE Spain), specifically in Osor. The ore vein, in both cases, is mainly composed of different sulfides (PbS, ZnS and also CaF2 in the case of Osor). During the activity period, the ore was extracted from the mines underground, and was concentrated using flotation techniques (Marques et al., 2003).

Fig. 1. Location of the studied mining districts: Val d'Aran (a) and Osor (b).

990


Since the mining closures (Val d'Aran in 1950 and Osor in 1980) the mines and the areas affected by the mining operations (including the mineral treatment factories) have not been reclaimed and, at present, the areas surrounding the factories are still covered by the remains of ore concentrates and wastes, which have become an important source of lead contamination (Pujals, 1992; Mata-Perelló, 1981). In addition, some spontaneous vegetation has started to re-forest the soils and mining landfills located in these areas. Among the higher plants found B. davidii, or the butterfly bush, which is an invasive plant of Chinese origin, was brought to Europe in the nineteenth century for use as a garden shrub. This plant spreads by producing abundant viable seeds and is now established in old quarries on masonry walls and in waste ground in both urban and rural areas (Mench et al., 2003). At present, this vegetation species has spread all over the mining districts studied. Mining wastes were sampled at three sites in the Val d'Aran and four sites at the Osor location. From the first mining district eleven samples were collected including two original Pb ore samples, eight samples of wastes from an old landfill with the remains of the Pb–Zn concentrates and one mine tailing from a mining dump. In Osor, seven samples were collected comprising two waste samples from a mining dump, two waste samples from a landfill with the remains of the Pb–Zn concentrates, one sample of the original lead ore and, finally, two soils probably contaminated by mining activities. For comparison purposes, five natural soils collected far away from the mining districts studied (in “La Garrotxa” region: 250 km from the Val d'Aran and 50 km from the Osor mining districts) and a leadedpetrol sample were also collected and analysed to determine the lead isotopic composition. Surface samples were taken using a polypropylene shovel, and subsequently transferred to clean polypropylene bags. Sampling of vegetation specimens was carried out in early autumn (just before shedding) to assure maximum metal accumulation. Leaves and flowering tops of B. davidii were sampled from the upper third part of the specimens and a composite sample was prepared for analysis from each sampling point. Control samples were collected in the surroundings of the Val d'Aran but far from the mining activities (on the opposite side of the Garona River, see Fig. 1), with the corresponding soil sample. All vegetation specimens were stored in clean polyethylene bags and kept in a plastic container to avoid contamination during transportation.

2.2. Sample treatment and procedures 2.2.1. Sample treatment In the laboratory, mine tailing and soil samples were oven-dried (50 °C), sieved (b 500 μm) and stored in polypropylene containers until analysis. Leaves and flowering tops of B. davidii specimens were washed thoroughly with deionised water to remove superficial dust and then oven-dried at 55 ± 5 °C for 24 h. To reduce particle size they were ground in an Agate ball mixer mill for 2– 5 min. Once plant tissues were powdered and dried, they were kept in capped polypropylene flasks until analysis. 2.2.2. Sample dissolution An ETHOS PLUS Milestone microwave with an HPR-1000/10 S high pressure rotor (Sorisole, Bergamo, Italy) was used for acid digestion of samples (mining wastes, soils and vegetation specimens). All the reagents used were of analytical grade Suprapur quality: nitric acid (Suprapur, Merck, Darmstadt, Germany) and hydrogen peroxide and hydrochloric acid (Trace Select, Fluka, Gilligham, Dorset, UK). Moreover, water obtained from a Milli-Q purifier system (Millipore Corp., Bedford, MA) was used throughout the study. For sample digestion of mine tailing and soil samples, about 500 mg of sample was weighed and placed in a PTFE reactor with 4 mL HNO3 (65%) and 12 mL HCl (37%). Then the reactor was sealed and heated following a two-stage digestion microwave program. The heating procedure consisted of a first stage of 5 min to reach 180 °C and a second stage of 10 min at 180 °C. After cooling, sample digests were filtered through a Whatman 42 filter, transferred into a 25 mL flask and brought to volume with MilliQ water. For vegetation specimens, 500 mg of sample was placed in a PTFE reactor with 9 mL HNO3 (65%) and 1 mL H2O2 (33%). The vessel was sealed and heated following the same two-stage digestion microwave programme described above. After cooling, sample digests were transferred into a 25 mL flask and brought to volume with MilliQ water. The lead contained in the leaded-petrol sample (10 mL) was extracted for a period of 24 h (rotatory shaking) with HNO3 (65%) using a 1:1 ratio. 2.2.3. Solution analyses A sequential inductively coupled plasma atomic emission spectrometer (ICP-AES, Liberty RL, Varian) was used to determine the lead content in studied samples. Standard solutions were prepared by serial dilution of a lead stock solution of 1000 ± 0.5 mg L− 1 (Pure Chemistry, ROMIL, UKAS calibration).


The performance of the experimental procedures was evaluated by analysing lead content in reference materials, including the sediment material BCR-701 from the Community Bureau of Reference (BCR, now the Standards, Measurements and Testing Programme) and the plant material GBW 07602 Bush Branches and Leaves from the National Research Centre for Certified Reference Materials, Beijing, China. In Table 1, certified and determined lead contents are presented. In both cases, good agreement was achieved between the certified/indicative values and the data determined. 2.3. Lead isotopic analysis A quadrupole-based ICP-MS system (Agilent 7500c, Agilent Technologies, Tokyo, Japan) equipped with an octapole collision reaction cell was used for the isotope ratio measurements. In this work, the collision/reaction cell was used only as an ion focussing lens and was therefore not filled with any pressurised gas. The operating conditions for ICP-QMS measurements were optimised daily by using a solution containing 10 μg L− 1 of Li, Tl, Y, Ce and Co and monitoring the intensities of the isotopes 205Tl, Y89, Li7, and their intensities at mass 156 (corresponding to 140 Ce16O+) and at mass 70 (corresponding to 140Ce2+) so as to monitor the percentage of doubly charged ions and of oxide ions, respectively. In addition, every working day, a 30 μg L− 1 lead solution (NIST SRM 981) was analysed in order to check the sensitivity and the resolution of lead isotope peaks to address peak tailing effects coming from adjacent peaks. Quality assurance of the isotope ratio measurements was achieved by testing the main parameters that affect the accuracy and precision of analytical data, including scan conditions, dead time, mass bias and interference effects. The lead isotope ratios 206Pb/207 Pb and 208Pb/ 207 Pb are employed in the present study as they are highly indicative of mining waste sample origin. In Table 2, the optimised parameters and instrumental bias corrections employed for lead isotopic determinations are summarised. Table 1 Certified and determined lead contents in BCR-701 and GBW 07602 reference materials

BCR-701 GBW 07602

Table 2 Optimised parameters and instrumental bias corrections employed for lead isotopic determinations RF power Gas flow rates Plasma Makeup Nebulizer Ion lenses Extract lens Einzel lens (1, 2 and 3) Isotopes measured Acquisition mode Replicate measurements Integration time for each isotope (s) Analysis time (min) Mass bias correction

Dead time (ns)

1500 W 15 L min− 1 0 L min− 1 1.17 L min− 1 3.2 V − 80, 10, − 80 V 206 Pb, 207Pb, 208Pb Three points per peak 5 5 6.25 Internal standard: Tl (203Tl/205Tl) Power law function Negligible (working range: 20–50 μg L− 1)

The suitability of the method developed for isotopic analysis was proved by analysing a 30 μg L− 1 lead solution of the isotopic standard NIST SRM 981 (National Institute of Standards and Technology, Gaithersburg, USA) spiked with a high purity Tl solution to 20 μg L− 1, using a Tl stock standard of 1000 ± 0.5 μg mL− 1 (Spectroscan, TECKNOLAB A/S, Norway). The measured, corrected and certified values for each isotope ratio studied are displayed in Table 3. As shown, the internal correction used in this study efficiently eliminates the error due to mass bias drift, increasing the accuracy. The overall precision of bias corrected Pb isotope ratios ranges from 0.16% to 0.18%. Analogous RSD values have been reported by other authors using similar instrumentation (Ettler et al., 2004; Xie and Kerrich, 2002). After measuring the Pb concentration in sample digests, the solutions were adjusted to a lead concentration of about 30 μg L− 1, according to the suitable Table 3 Comparison of measured, corrected (thallium internal correction) and certified lead isotope ratios using a 30 μg L− 1 NIST SRM 981 solution spiked with 20 μg L− 1 of thallium (n = 5, 95% confidence level)

Certified

Determined

206

454 ± 19 7.1 ± 0.7

446 ± 6 6.5 ± 0.9

208

Determined results are expressed as the mean of four parallel sample determinations with the corresponding statistical confidence range (95% confidence level).

991

Pb/207Pb Pb/207Pb

Measured values

Corrected values

Certified values

1.100 ± 0.001 (+0.6%) 2.360 ± 0.003 (− 0.4%)

1.095 ± 0.002 (+0.1%) 2.370 ± 0.004 (− 0.01%)

1.0933 ± 0.0003 2.3704 ± 0.0005

In brackets are displayed error percentages between obtained and certified values.

992


working range (from 20 to 50 μg L− 1) selected previously (see Table 2). Once dilutions were made, they were spiked with a Tl high purity solution to 20 μg L− 1 to correct for mass bias drift. For each sample, the two independent pre-treated replicates were separately analysed five times. The mean value and respective confidence range (95% confidence level) were then calculated. The measurement procedure was evaluated by analysing a 30 μg L− 1 NIST SRM 981 solution at the beginning and at the end of each batch of samples. 2.4. Statistical analysis The statistical analysis of the obtained results was performed using the program SPSS 12.0 for Windows®. Cluster analysis (Norusis, 1993) was employed as a statistical procedure to form groups of similar objects (samples) according to their isotopic composition. Correlation was measured between the isotope ratios using Pearson's correlation coefficient. 3. Results and discussion 3.1. Lead content in samples Remarkable differences in lead burdens were observed in the different wastes collected due to the different nature of the materials (Table 4). The highest concentrations were found in the remains of the Pb–Zn concentrates with values up to 13% of Pb. In general, the mine tailings from the Val d'Aran are substantially more enriched in lead than those collected in the mining district of Osor. In both cases, the lead concentrations determined are of the same order as those reported for similar mining areas (Ye et al., 2002; Mackenzie and Pulford, 2002). On the other hand, the data revealed that the soils collected at Osor exhibited lead contents of the same order of magnitude as wastes from the mining dump. From these results and, taking into account the location of the soils, it was therefore reasonable to assume that the high lead amounts were derived principally from the mining operations. As commented in the Experimental section, five natural soils were also collected for comparison purposes and the average of lead concentration recorded was 13 mg kg− 1. The lead concentration found in the control soil was significantly higher, with a mean value of 158 mg kg− 1. However, both concentrations are within the range of lead content for surface soils in different countries reported in the literature (Kabata-Pendias, 2001).

Table 4 Lead concentrations (mg kg− 1) and lead isotope ratios in mining tailings and soil samples

Val d'Aran mining district Original Pb ore (PbS) Ore(VA)-1 (Victoria mine) Ore(VA)-2 (Liat mine) Mining Dump (MD) MD(VA)-1 Concentrate remains (CR) CR(VA)-1 CR(VA)-2 CR(VA)-3 CR(VA)-4 CR(VA)-5 CR(VA)-6 CR(VA)-7 CR(VA)-8 Osor mining district Original Pb ore (PbS) Ore(O)-1 Mining dump (MD) MD(O)-1 MD(O)-2 Concentrate remains (CR) CR(O)-1 CR(O)-2 Contaminated soils (CS) CS(O)-1 CS(O)-2 Natural soils (NS) NS-1 NS-2 NS-3 NS-4 NS-5 Control soil (SS)

Pb (mg kg− 1) a

206/207

860,000– 870,000 c 860,000– 870,000 c

1.156 ± 0.003

2.440 ± 0.006

1.153 ± 0.004

2.436 ± 0.009

20,000 ± 1000

1.155 ± 0.007

2.439 ± 0.013

34,300 ± 200 60,900 ± 500 79,700 ± 600 130,000 ± 4000 13,100 ± 200 8300 ± 200 12,200 ± 300 66,400 ± 600

1.158 ± 0.005 1.157 ± 0.004 1.159 ± 0.004 1.159 ± 0.004 1.159 ± 0.006 1.159 ± 0.004 1.156 ± 0.003 1.157 ± 0.004

2.4380 ± 0.010 2.4380 ± 0.009 2.4405 ± 0.007 2.4408 ± 0.009 2.4396 ± 0.007 2.4359 ± 0.007 2.4386 ± 0.007 2.4407 ± 0.007

330,000 ± 3000

1.186 ± 0.008

2.481 ± 0.010

320 ± 10 2050 ± 40

1.185 ± 0.006 1.185 ± 0.006

2.480 ± 0.010 2.479 ± 0.011

28,000 ± 600 38,970 ± 20

1.183 ± 0.005 1.182 ± 0.009

2.479 ± 0.012 2.484 ± 0.017

1200 ± 100 770 ± 10

1.183 ± 0.006 1.184 ± 0.005

2.477 ± 0.014 2.483 ± 0.010

10 ± 0.8 15.0 ± 0.5 13 ± 0.3 10.0 ± 0.7 15 ± 1 158 ± 4

1.173 ± 0.003 1.172 ± 0.003 1.172 ± 0.002 1.177 ± 0.003 1.178 ± 0.004 1.161 ± 0.003

2.452 ± 0.007 2.4526 ± 0.005 2.4515 ± 0.005 2.4539 ± 0.006 2.4691 ± 0.008 2.4370 ± 0.006

Pb b

208/207

Pb b

a Results are expressed as mean of duplicate samples with the corresponding standard deviation (±SD). b Results are expressed as mean of two independent pre-treated replicates analysed five times separately with the respective confidence range (95% confidence level). c Data obtained from (Pujals, 1992).

Regarding vegetation specimens, considerable lead contents were found in B. davidii specimens growing on the mine tailings of the Val d'Aran area (Table 5). Comparing the metal content of vegetation samples from

E. Marguí et al. / Science of the Total Environment 367 (2006) 988–998 Table 5 Lead concentrations (mg kg− 1) and lead isotope ratios in Buddleia davidii specimens (L = leaves and F = flowering tops) Pb (mg kg− 1) a

206/207

Pb b

208/207

Pb b

Val d'Aran mining district L(VA)-1 47.5 ± 0.1 L(VA)-2 607 ± 2 L(VA)-3 300 ± 10 L(VA)-4 277 ± 7 L(VA)-5 84 ± 1 L(VA)-6 59.3 ± 0.4 L(VA)-7 68 ± 2 L(VA)-8 135.6 ± 0.2 F(VA)-1 86 ± 2 F(VA)-2 360 ± 10 F(VA)-3 430 ± 30 F(VA)-4 300 ± 10 F(VA)-5 175 ± 4 F(VA)-6 650 ± 20

1.154 ± 0.006 1.158 ± 0.007 1.156 ± 0.005 1.154 ± 0.007 1.153 ± 0.004 1.154 ± 0.005 1.156 ± 0.007 1.156 ± 0.003 1.155 ± 0.006 1.156 ± 0.007 1.156 ± 0.005 1.157 ± 0.005 1.155 ± 0.006 1.154 ± 0.006

2.439 ± 0.012 2.441 ± 0.013 2.438 ± 0.011 2.439 ± 0.015 2.440 ± 0.008 2.440 ± 0.010 2.436 ± 0.013 2.438 ± 0.009 2.441 ± 0.012 2.440 ± 0.012 2.442 ± 0.012 2.441 ± 0.011 2.441 ± 0.011 2.439 ± 0.012

Osor mining district L(O)-1 26 ± 2 L(O)-2 45.45 ± 0.01 L(O)-3 12.4 ± 0.4 F(O)-1 10.8 ± 0.1 F(O)-2 40 ± 2 F(O)-3 7.3 ± 0.4

1.183 ± 0.005 1.185 ± 0.009 1.184 ± 0.006 1.184 ± 0.004 1.184 ± 0.005 1.182 ± 0.005

2.481 ± 0.011 2.480 ± 0.017 2.476 ± 0.011 2.478 ± 0.008 2.476 ± 0.010 2.478 ± 0.011

Control sample L 5.7 ± 0.5 F 8.1 ± 0.1

1.169 ± 0.004 1.172 ± 0.004

2.446 ± 0.008 2.445 ± 0.009

a

Results are expressed as mean of duplicate samples with the corresponding standard deviation (± SD). b Results are expressed as mean of two independent pre-treated replicates analysed five times separately with the respective confidence range (95% confidence level).

the mining districts with the corresponding metal content in the control sample using the Relative Absorption Coefficient (RAC: content in plant growing on mining affected area/content in control sample), it can be concluded that an accumulation process has occurred. Relative absorption coefficients reach values up to 106 and 80 for leaves and flowering tops, respectively. In order to identify if the anomalous lead content in the vegetation specimens is mainly derived from mining activities rather than from other possible lead sources including past leaded gasoline emissions, the Pb isotopic composition of plants and wastes was determined and a comparison study was performed. 3.2. Lead isotope ratio analysis Results for 206Pb/207Pb ratios versus 208 Pb/207Pb ratios obtained in the isotopic analysis of mine tailings, soils and the leaded-petrol sample are plotted in Fig. 2a.

993

As can be seen, a correlation between the pair of ratios exists. As suggested in the literature, the presence of a straight line appears to represent the simple mixing of two dominant components, known as end members, which would indicate different sources of lead (Bacon et al., 1996; Bacon, 2002). This fact could be explained by taking into account the possible variations of the relative abundance of lead isotopes over time, as commented in the Introduction section. Three probable sources of lead contamination in vegetation specimens collected at the two mining areas are considered: past leaded-petrol emissions, natural background lead and, obviously, the lead contained in the ores from the mining operations. At a cursory glance, as shown in Fig. 2a, mine tailing samples, natural soil samples and the leaded-petrol sample fall into clearly distinct groups corresponding to the different lead sources pointed out previously. It is interesting to note that the measured changes in isotope ratios are much greater than the uncertainties in the measurements and that they are significant in all cases. This trend highlights that the implemented ICP-QMS method is precise enough to discriminate among different lead origins. Besides, it is also possible to differentiate the two mining districts studied. All the mine tailings from the Val d'Aran can be assembled in a single group with Pb ratios in the range 1.153 to 1.159 (206Pb/207Pb) and 2.436 to 2.441 (208Pb/207Pb). In a similar way, the samples collected at Osor are grouped in values ranging from 1.182 to 1.186 (206Pb/207Pb) and from 2.447 to 2.484 (208 Pb/207Pb). However, the Pb isotopic composition of the different kinds of samples within a mining district (including galena ore samples, Pb–Zn concentrate remains and tailings from the mining dumps) was indistinguishable. On the other hand, the resemblance between the composition of likely contaminated soil samples collected at Osor and those from mine tailings point to mining activities as the source of soil contamination. The isotopic composition of the vegetation specimens has also been included in Fig. 2b. The obtained data set shows that there was no noticeable contamination of B. davidii specimens from petrol emissions since lead isotope ratios determined in the samples are substantially different from those obtained in the leadedpetrol sample analysed, for which a value as low as 1.080 for 206 Pb/207Pb ratio was obtained, in agreement with data reported in the literature (Charalampides and Manoliadis, 2002; Bellis et al., 2004). The reduction in the use of alkyl-Pb additives in gasoline during the past years has led to a gradual increase of the 206 Pb/207Pb ratio in environmental samples affected by atmospheric

994


(a) 2,520 2,500

208Pb/207Pb

2,480 2,460

Osor mining district

Mining tailings Petrol Natural soils Control soil

2,440 2,420 Val d'Aran mining district

2,400 2,380 2,360

R2 = 0,9539

2,340 2,320 1,060

1,080

1,100

1,120

1,140

1,160

1,180

1,200

1,180

1,200

206Pb/207Pb

(b)

208Pb/207Pb

2,520 2,500

Mining tailings

2,480

Petrol

2,460

Natural soils

2,440

vegetation specimens

2,420

control soil

2,400 2,380 2,360 2,340 2,320 1,060

1,080

1,100

1,120

1,140

1,160

206Pb/207Pb

(c) 2,520 Mining tailings 2,500

208Pb/207Pb

2,480 2,460

Osor mining district

Natural soils vegetation specimens Control soil

2,440 2,420 2,400 2,380 1,140

Control sample Val d'Aran mining district

1,150

1,160

1,170

1,180

1,190

1,200

206Pb/207Pb

Fig. 2. Plot of 206Pb/207Pb ratios versus 208Pb/207Pb for all samples studied. Results are represented as the mean of two independent pre-treated replicates analysed five times separately, with the corresponding statistical confidence range (95% confidence level). a: isotopic composition of samples corresponding to sources of lead contamination, b: results for vegetation samples are included, c: enlarged view of the plot by excluding petrol sample.

deposition. For example, the 206Pb/207 Pb ratio in moss samples collected at different localities in Norway have risen from 1.120 (1977) to 1.143 (2000) (Steinnes et al., 2005). However, the ratios determined in the vegetation specimens studied in the present work are substantially

higher (from 1.154 to 1.183) than current atmospheric deposition values. Moreover, as shown in Fig. 2c, the lead isotopic composition of B. davidii samples collected at the mining areas is close to that determined for the mine


995

207

tailings from the same areas, suggesting that the anomalous amounts of Pb in the plants are derived from mining activities rather than from other lead sources. It is interesting to note that the results for the control sample do not match with the groups formed for the vegetation specimens collected at the Osor and Val d'Aran mining districts. On the contrary, the isotopic composition of this sample is similar to those measured for the natural soils studied, which are representative of the estimated natural background composition. Therefore, it seems feasible to discriminate between the control sample and lead contaminated specimens taking into account only the values of 206Pb/207Pb and 208Pb/

Pb isotope ratios. This trend is more clearly shown in the dendrogram obtained after performing a cluster analysis of the isotope ratios of vegetation specimens and mine tailings, using the average linkage between groups (Fig. 3). Three distinct groups, corresponding to the samples collected at the Val d'Aran mining district, the samples from Osor and the control sample, can be observed. In order to gain more complete information about the similarity between lead isotope ratios from B. davidii tissues (leaves and flowering tops) and those from the mine tailings or soil samples where the plants are growing, the 206Pb/207Pb values obtained in both environmental matrices are compared in Fig. 4. The results Rescaled Distance Cluster Combine

0

5

10

Flowering tops(2) Leaves(3) Contaminated Soil Flowering tops(1) Flowering tops(3) Concentrate Remains Leaves(1) Mining Dump Mining Dump Leaves(2) Ore Vein Concentrate Remains Contaminated Soil Control sample (Flowering tops) Control sample (Leaves) Concentrate Remains Flowering tops(7) Flowering tops(2) Concentrate Remains Concentrate Remains Concentrate Remains Leaves(5) Leaves(6) Leaves(3) Concentrate Remains Concentrate Remains Concentrate Remains Leaves(10) Leaves(8) Leaves(9) Mining Dump Flowering tops(6) Leaves(4) Flowering tops(5) Leaves(1) Flowering tops(1) Flowering tops(4) Ore Vein Leaves(2) Flowering tops(3) Concentrate Remains Ore Vein

Fig. 3. Dendrogram resulting from Cluster analysis of the average linkage between groups (n = 42).

206

Pb/207Pb and

15

20

25

OSOR

VAL D'ARAN

208

Pb/207Pb ratios of vegetation specimens and mining tailings, using

996

E. Marguí et al. / Science of the Total Environment 367 (2006) 988–998 Tailing/Soil

Leaves

CR (VA)

CR (VA)

Flowering

1,200 1,195 1,190

Pb(206/207)

1,185 1,180 1,175 1,170 1,165 1,160 1,155 1,150 1,145 1,140 SS

CR (VA)

CR (VA)

CR(O)

CS(O)

Fig. 4. Comparison of 206Pb/207Pb ratios in mining tailings, soil samples and corresponding vegetation specimens. Results are represented as the mean of two independent pre-treated replicates analysed five times separately, with the corresponding statistical confidence range (95% confidence level). SS: Control soil, CR: remains of Pb–Zn concentrates, CS: contaminated soil, (VA): Val d'Aran, (O): Osor.

show that 206 Pb/207Pb ratios in mine tailings and the corresponding B. davidii samples growing on them were identical. A different trend was observed in the case of the control sample, which presents a 206Pb/207Pb ratio different from that of the corresponding soil sample, indicating that in this case the lead uptake from the soil did not play a major role. This fact could be explained if one considers the high mobility of lead contained in the mining wastes. As pointed out in a previous work (Marguí et al., in press), the oxidation of the ore vein (PbS) into more mobile species (PbSO4) frequently occurs and, in such a case, the lead could be more easily taken up by vegetation specimens. Moreover, it has to be

Before cleaning

kept in mind that the acidic pH of mining waste samples (from 3.8 to 6.6) also facilitates the metal availability to vegetation specimens as has been addressed in previous works (Li et al., 1998). Furthermore, in order to study the possible presence of lead in the surface of leaves and flowering tops of B. davidii, an analysis of extracted water soluble metals was performed. For this purpose, 1 g of intact vegetation specimen was immersed in 10 mL of deionised water for 10 min. The lead concentrations analysed in the water extracts were, in most cases, below 20 μg L− 1. The 206Pb/ 207 Pb values obtained in the analysis of the leaves and flowering tops of B. davidii before and after water

After cleaning

Water extract

1,200 1,190

Pb(206/207)

1,180 1,170 1,160 1,150 1,140 1,130 1,120 1L(O)

1F(O)

2L(O)

2F(O)

3L(O)

3F(O) 1L(VA) 1F(VA) 2L(VA) 2F(VA)

Fig. 5. Graphical representation of 206Pb/207Pb ratios obtained before cleaning, after cleaning and in water extracts of Buddleia davidii specimens (isotopic composition of water extracts was carried out whenever [Pb] N 20 μg L− 1, according to the suitable working range: 20–50 μg L− 1). Results are represented as the mean of two independent pre-treated replicates analysed five times separately, with the corresponding statistical confidence range (95% confidence level). (VA): Val d'Aran, (O): Osor, F: flowering tops of Buddleia davidii, L: leaves of Buddleia davidii.


cleaning were indistinguishable and were also identical to those from the isotopic analysis of water extracts (Fig. 5). To sum up, the above points indicate that mining activity could be the unique source of lead contamination in the B. davidii specimens studied. 4. Conclusions Elevated concentrations of Pb were found in the different mine tailings collected from the Val d'Aran and Osor mining districts, indicating a potential environmental hazard. Similarly, B. davidii specimens growing in these areas presented a higher lead content compared to the natural burden which is characteristic of this vegetation species (control sample). The use of isotopic composition to identify the origin of the anomalous lead content in vegetation specimens proved to be a powerful technique. The precision and accuracy of ICP-QMS analysis made it possible to satisfactorily discriminate among the likeliest lead sources in the studied areas. The results show that the lead in the leaves and flowering tops of B. davidii specimens from the mining areas did not have the same isotopic composition as the specimen from an uncontaminated site (control sample) but was close to that of the mining tailings collected. This trend suggests that the lead accumulation in plants is primarily derived from the mining operations rather than from other lead sources such as leaded-petrol emissions. It was concluded that the abandoned mining areas studied (particularly the Val d'Aran mining district) can become important sources of lead contamination for the vegetation growing in the soils and on the mining landfills located in these areas. Acknowledgements The authors express their sincere gratitude to Giovanni Pardini (Soil Science Unit, University of Girona) for his help during the sampling campaign in the Osor mining district. This study was financed by the Spanish National Research Programme (CGL200405963-C04-03/HID). E. Marguí gratefully acknowledges a grant from the Autonomous Government of Catalonia (Ref.2002FI 00577). References Al-Ammar AS, Barnes RM. Improving isotope ratio precision in inductively coupled plasma quadrupole mass spectrometry by common analyte internal standardization. J Anal At Spectrom 2001;16:327–32.

997

Bacon JR. Isotopic characterisation of lead deposited 1989–2001 at two upland Scottish locations. J Environ Monit 2002;4:291–9. Bacon JR, Dinev NS. Isotopic characterisation of lead in contaminated soils from the vicinity of a non-ferrous metal smelter near Ploudiv, Bulgaria. Environ Pollut 2005;134:247–55. Bacon JR, Jones KC, McGrath SP, Johnston AE. Isotopic character of lead deposited from the atmosphere at a Grassland site in the United Kingdom since 1860. Environ Sci Technol 1996;30:2511–8. Bellis DJ, Satake K, McLeod CW. A comparison of lead isotope ratios in the bark pockets and annual rings of two beech trees collected in Derbyshire and South Yorkshire, UK. Sci Total Environ 2004;321:105–13. Boulyga SF, Testa CT, Desideri D, Becker JS. Optimisation and application of ICP-MS and alpha-spectrometry for determination of isotopic ratios of depleted uranium and plutonium in samples collected in Kosovo. J Anal At Spectrom 2001;16:1283–9. Charalampides G, Manoliadis O. Sr and Pb isotopes as environmental indicators in environmental studies. Environ Int 2002;28:147–51. Cheng Z, Foland KA. Lead isotope in tap water: implications for Pb sources within a municipal water supply system. Appl Geochem 2005;20:353–65. Elbering B, Asmund G, Kunzendorf H, Krogstad EJ. Geochemical trends in metal-contaminated fiord sediments near a former lead– zinc mine in west Greenland. Appl Geochem 2002;17:493–502. Ettler V, Mihaljevic M, Komarek M. ICP-MS measurements of lead isotopic ratios in soils heavily contaminated by lead smelting: tracing the sources of pollution. Anal Bioanal Chem 2004;378:311–7. Jackson BP, Winger PV, Lasier PJ. Atmospheric lead deposition to Okefenokee Swamp, Georgia, USA. Environ Pollut 2004;130:445–51. Kabata-Pendias A. Trace elements in soils and plants. Third edition. CRC press LLC; 2001. Kurkjian R, Dunlap C, Flegal AR. Long-range downstream effects of urban runoff and acid mine drainage in the Debel River, Armenia: insights from lead isotope modelling. Appl Geochem 2004;19:1567–80. Labonne M, Othman DB, Luck J-M. Pb isotopes in mussels as tracers of metal sources and water movements in a lagoon (Than Basin, S. France). Chem Geol 2001;181:181–91. Li FL, Shan XQ, Zhang SZ. Evaluation of plant availability of rare earth elements in soils by chemical fractionation and multiple regression analysis. Environ Pollut 1998;102:269–77. Mackenzie AB, Pulford ID. Investigation of contaminant metal dispersal from a disused mine site at Tyndrum, Scotland, using concentration gradients and stable Pb isotope ratios. Appl Geochem 2002;17:1093–103. Marguí, E., Queralt, I., Carvalho M.L., Hidalgo, M. Assessment of metal availability to vegetation (Betula pendula) in Pb–Zn ore concentrate residues with different features. Environ Pollut 2006, in press. Marques AF, Queralt I, Carvalho ML, Bordalo M. Total reflection Xray fluorescence and energy-dispersive X-ray fluorescence analysis of runoff water and vegetation from abandoned mining of Pb–Zn ores. Spectro Chim Acta, Part B: Atom Spectrosc 2003;58:2191–8. Mata-Perelló J.M., “Els minerals a Catalunya”. Doctoral Thesis. Faculty of Geology. University of Barcelona, Spain (1981). Mench M, Bussière S, Boisson J, Casting E, Vangronsveld J, Ruttens A, et al. Progress in remediation and revegetation of the barren Jales gold mine spoil after in situ treatments. Plant Soil 2003;249:187–202. Négrel P, Roy S. Investigating the sources of the labile fraction in sediments from silicate-drained rocks using trace elements, and strontium and lead isotopes. Sci Total Environ 2002;298:163–81.

998


Négrel P, Kloppmann W, Garcin M, Giot D. Lead isotope signatures of Holocene fluvial sediments from the Loire River valley. Appl Geochem 2004;19:957–72. Norusis M.J., SPSS® for Windows™, Professional statistics™, Release 6.0. SPSS Inc. (1993). Platzner IT. Modern isotope ratio mass spectrometry. Wiley, England: John Wiley & Sons; 1997. Prego R, Cobelo-García A. Cadmium, copper and lead contamination of the seawater column on the Prestige shipwreck (NE Atlantic Ocean). Anal Chim Acta 2004;524:23–6. Pujals I., “Las mineralizaciones de sulfuros en el cambro-ordovícico de la Val d'Aran (Pirineo Central, Lérida)”. Doctoral Thesis. Faculty of Sciences. Autonomous University of Barcelona, Spain, 1992. Rabinowitz MR. Plant uptake of soil and atmospheric lead in southern California. Chemosphere 1972;4:175–83. Spiro B, Weiss DJ, Purvis OW, Mikhailova I, Williamson BJ, Coles BJ, et al. Lead isotopes in lichen transplants around a Cu smelter in

Russia determined by MC-ICP-MS reveal transient records of multiple sources. Environ Sci Technol 2004;38:6522–8. Steinnes E, Åberg G, Hjelmseth H. Atmospheric deposition of lead in Norway: spatial and temporal variation in isotopic composition. Sci Total Environ 2005;336:105–17. Vinagre C, França S, Costa MJ, Cabral HN. Accumulation of heavy metals by flounder, Platichthys flesus (Linnaeus 1758), in a heterogeneously contaminated nursery area. Mar Pollut Bull 2004;49:1109–13. Xie Q, Kerrich R. Isotope ratio measurement by hexapole ICP-MS: mass bias effect, precision and accuracy. J Anal At Spectrom 2002;17:69–74. Ye ZM, Shu WS, Zhang ZQ, Lan CY, Wong MH. Evaluation of major constraints to revegetation lead/zinc mine tailings using bioassay techniques. Chemosphere 2002;47:1103–11.