Journal of Geochemical Exploration xxx (2011) xxx–xxx

GEXPLO-04924; No of Pages 8

Contents lists available at SciVerse ScienceDirect

Journal of Geochemical Exploration

j ourna l homepage: www.e lsev ie r .com/ locate / jgeoexp

Differences in the bioaccessibility of metals/metalloids in soils from mining andsmelting areas (Copperbelt, Zambia)

Vojtěch Ettler a,⁎, Bohdan Kříbek b, Vladimír Majer b, Ilja Knésl b, Martin Mihaljevič a

a Institute of Geochemistry, Mineralogy and Mineral Resources, Charles University in Prague, Faculty of Science, Albertov 6, 128 43 Praha 2, Czech Republicb Czech Geological Survey, Geologická 6, 152 00 Praha 5, Czech Republic

⁎ Corresponding author. Tel.: +420 221 951 493; faxE-mail address: ettler@natur.cuni.cz (V. Ettler).

0375-6742/$ – see front matter © 2011 Elsevier B.V. Alldoi:10.1016/j.gexplo.2011.08.001

Please cite this article as: Ettler, V., et al., D(Copperbelt, Zambia), J. Geochem. Explor.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 September 2010Accepted 13 August 2011Available online xxxx

Keywords:BioaccessibilityTopsoilMetalsArsenicCopperbeltZambia

Differences in the total and bioaccessible concentrations of As and metals (Co, Cu, Pb, Zn) in topsoils(n=107) from the mining and smelting areas in the Zambian Copperbelt were evaluated. The mean totalconcentrations of metals and As in topsoils were generally 2 to 7× higher in the smelting area, indicating sig-nificantly higher effect of smelter dust fallout on the degree of topsoil contamination. The contaminant bioac-cessibility was tested by an US EPA-adopted in vitro method using a simulating gastric fluid containing a0.4 M solution of glycine adjusted to pH 1.5 by HCl. Higher bioaccessibilities in the smelter area were ob-served for As and Pb, attaining 100% of the total metal/metalloid concentration. The maximum bioaccessibil-ities of As and Pb in the mining area were 84% and 81%, respectively. The ranges, mean and medianbioaccessibilities of Co, Cu and Zn were similar for the two areas. The maximum bioaccessibilities of Co, Cuand Zn were 58–65%, 80–83% and 79–83%, respectively. The obtained data indicate that a severe healthrisk related to topsoil ingestion should be taken into account, especially in smelting areas.

: +420 221 951 496.

rights reserved.

ifferences in the bioaccessibility of metals/m(2011), doi:10.1016/j.gexplo.2011.08.001

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Themining and smelting activities are responsible for extensive con-tamination of soils. The smelter emissions as well as wind-blown dustfrom mine tailings and smelter slag dumps are generally the mainpoint sources of soil pollution (Ettler et al., 2005a, 2009, 2011; Kříbeket al., 2010; Šráček et al., 2010; Vítková et al., 2010). Studies dealingwith the bioavailability and bioaccessibility of metals/metalloids con-taminants in highly-polluted soils are extremely useful in understand-ing the possible effect on biota (Bosso and Enzweiler, 2008; Chen etal., 2009; Douay et al., 2008; Juhasz et al., 2011; Roussel et al., 2010).In particular, human exposure to contaminants in mining/smeltingareas has implications for health risk assessment (Banza et al., 2009;Roussel et al., 2010).

The “bioaccessible” fraction is defined as the amount of contaminantthat is mobilized from the solid matrix (e.g. soil) in the human gastroin-testinal tract and becomes available for intestinal absorption. The “bio-available” fraction is the fraction of contaminant that can reach theblood stream from the gastrointestinal tract (Morrison and Gulson,2007; Roussel et al., 2010; Ruby et al., 1999). In the last two decades, anumber of laboratory methods (often called PBET, physiologically-based extraction tests) have been developed to investigate in vitro theoral (ingestion) or respiratory bioavailability/bioaccessibility of metalsfrom polluted geomaterials (soils, wastes) (Oomen et al., 2002, 2003a,

2003b; Ruby et al., 1993; Schroder et al., 2004). These methods andtheir applications have recently been reviewed by Plumlee and Ziegler(2006) and Plumlee et al. (2006) and have led to the development ofstandardized tests adopted by theU.S. Environmental ProtectionAgency(US EPA, 2007). Although this test was validated by in vivo tests only forPb and As (Ruby et al., 1993, 1996; Schroder et al., 2004), it has also beenwidely adopted to study the bioaccessibility of other inorganic contam-inants in polluted soils (e.g., Kim et al., 2002; Madrid et al., 2008a,2008b).

The present study is based on our previous screening soil surveydiscriminating the contaminant sources in the area of intense copper–cobalt mining and smelting in the Zambian Copperbelt (Kříbek et al.,2010). It has been reported that children can ingest between tens andhundreds of milligrams of soil per day via hand-to-mouth behaviour.Up to 200 mg soil/day was observed by van Wijnen et al. (1990) and,for the 90th percentile, typically between 40 and 100 mg/day. Morerecently, Özkaynak et al. (2011) used a USEPA Stochastic HumanExposure and Dose Simulation Model (SHEDS) to show that up to1367 mg soil/day can be ingested with a 95th percentile of 176 mg/dayand mean value of 41 mg/day. Thus, a severe risk of exposure to metalliccontaminants in highly polluted areas of the Zambian Copperbeltcan be anticipated. High exposure to metal contaminants expressedparticularly as high urinary Co concentrations was also reportedfrom the nearby Copperbelt mining and smelting district in theDemocratic Republic of Congo (Banza et al., 2009). As a result, thisstudy is focused on investigation of the differences in gastric bioac-cessibility of metals (Co, Cu, Pb, Zn) and As in topsoils from two dis-tinct areas with contrasting pollution sources (mining vs. smelting).

etalloids in soils from mining and smelting areas

2 V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

2. Materials and methods

2.1. Soil sampling

Based on extensive data on the spatial distribution of inorganic con-taminants in the Zambian Copperbelt district (Kříbek et al., 2010), twohot-spots with contrasting sources of pollution were selected for inves-tigation of the metal/metalloid bioaccessibility: 1) a mining area in thevicinity of Chingola with a number of active open-pit mines (Nchangaand Chingola) (n=52 soil samples) and 2) a smelting area in the vicin-ity of Kitwe with the Nkana Cu smelter active between 1932 and 2009(n=55 soil samples) (Fig. 1). According to Mihaljevič et al. (2010),the prevailing wind direction in the studied areas is NE–SW betweenNovember and February (wind speed up to 2 m/s) whereas, duringthe rest of the year, stronger winds with a velocity of N3 m/s in the di-rection SE-NW prevail. The wind direction has a significant effect onthe spatial distribution of airborne contamination in the vicinity ofpoint pollution sources in the Zambian Copperbelt (Ettler et al., 2011;Kříbek et al., 2010; Mihaljevič et al., 2010). In the mining area, thedust fallout originates mainly from open-pit mining operations, orecrushers, ore/concentrate transport and mine tailings. In contrast, theareas around smelters are mainly affected by the smelter emissionsand fine-grained slag dust generated by slag treatment plants (crushingprior to further re-smelting and further metal recovery) (Kříbek et al.,2010; Vítková et al., 2010).

Only topsoil samples (0–2 cm depth) were considered in thisstudy, being the most probable source of potential health risk dueto ingestion. According to Soil Taxonomy (Soil Survey Staff, 2010),the soils were characterized as Oxisols. The samples were stored inpolyethylene (PE) bags, air-dried to constant weight on returning tothe laboratory and sieved through a clean 0.25-mm stainless steelsieve (Retsch, Germany). The 0.25-mm sieved fraction was used forthe pH determination and bioaccessibility testing, because this parti-cle size is representative of that which adheres to children's hands

Kafue R

Chingola

Kitwe (Nkana)

Fig. 1. The map of the Zambian Copperbelt

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

(US EPA, 2007). An aliquot part of each sample was finely ground inan agate mortar (Fritsch Pulverisette, Germany) and used for subse-quent bulk chemical analysis.

2.2. Soil analysis

The pH measurements were performed according to Pansu andGautheyrou (2006) in a 1:5 (w/v) soil-deionized water suspensionafter 1-h agitation using a Schott Handylab pH meter. Total organiccarbon (Corg) and total inorganic carbon (Ccarb) contents were deter-mined using Eltra CS 500 analyzer (Eltra, Germany). Total sulphur(Stot) was determined on Eltra CS 530 analyzer (Eltra, Germany).

The pseudo-total digests of soil samples were obtained by a stan-dardized aqua regia extraction protocol according to ISO Standard11466 (ISO, 1995). Certified reference material (CRM) BCR-483 (sew-age sludge-amended soil) and standard reference material (SRM)NIST 2711 (Montana soil) were used to control the accuracy of theaqua regia pseudo-total digestion, yielding satisfactory values (Table 1).Although NIST 2711 has element values certified for total digests, theaqua regia pseudo-total digests were in good agreement with the certi-fied values as well as with the aqua regia data recently published forthis SRM (Karadaş and Kara, 2011). Total digests were analyzed for thecontent of Co, Cu, Pb and Znby a Perkin Elmer 4000flame atomic absorp-tion spectrometer (FAAS) or by a Thermo Scientific Xseries 2 inductivelycoupled plasma mass spectrometer (ICP-MS). The As concentrationswere determined by a Perkin Elmer 503 hydride generation atomic ab-sorption spectrometer (HG-AAS) or by ICP-MS.

The bioaccessibility test was performed according to the US EPA(2007) protocol, identical with the Simple Bioaccessibility ExtractionTest (SBET) adopted by the British Geological Survey (Oomen et al.,2002). The extraction fluid contained 0.4 M glycine (30.028 g glycinedissolved in 800 ml of deionizedwater), adjusted to pH1.5±0.05by re-agent grade HCl (Merck, Germany), finalized by diluting to 1 l by deio-nized water (MilliQ+, Millipore Academic, USA) and pH verification. A

50 km

iver

ZAMBIA

ZIMBABWE

MA

LAW

I

DR CONGO

AN

GO

LA

TANZANIA

MOZAMBIQUECopperbeltProvince

24 Eo 25 Eo

12 So

studyarea

N

smeltermines

location and study area (dashed line).

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

Table 1Quality control of the aqua regia pseudo-total digestion (mean±standard deviation).

Code As (ppm) Co (ppm) Cu (ppm) Pb (ppm) Zn (ppm)

BCR 483 (n=1)measured –a – 353 554 1014certified – – 362±12 501±47 987±37

NIST 2711 (n=3)measured 91.2±1.6 7.5±0.4 104±28 1118±25 335±7certifiedb 105±8 10c 114±2 1162±31 350.4±4.8

a –, not given.b Certified for total content, not for aqua regia pseudo-total digestion.c Noncertified value (for information only).

3V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

solid-to-fluid ratio of 1/100 was used for the extraction. A mass corre-sponding to 0.5 g of sieved soil sample was placed in 100-ml high-density polyethylene (HDPE) bottles (P-lab, Czech Republic), 50 ml ofextraction fluid was added and the mixture was agitated for 2 h at37 °C. After the extraction procedure, the extract was filtered through0.45-μm nitrocellulose membrane filters (Millipore, USA), diluted andanalyzed for the total contents of As, Co, Cu, Pb and Zn by HG-AAS,FAAS or ICP-MS. The bioaccessible concentrations of metals and Aswere expressed in mg/kg (ppm) and converted to % amount of totalcontent. The extractionwas performed in triplicate for ten randomly se-lected samples and indicated that the reproducibility of the procedurewas generally below10%, but never exceeded 20% RSD (higher standarddeviations were observed for some samples with bioaccessible concen-trations below 5 ppm). The bioaccessibility test employed, simulatinggastric conditions with low pH, is a suitable predictor for estimation ofthe “worst case” situation for physiologically relevant fasting conditions(Oomen et al., 2002; Ruby et al., 1993).

2.3. Data treatment

The basic statistics of the obtained data were calculated by Excel2003 (MS Office, Microsoft, USA). The grid was calculated and the re-sults of the spatial distribution of metal/metalloid contaminants (bulkconcentrations and bioaccessibility data) were mapped using Surfer8 (Golden Software, USA). The correlation coefficients were calculat-ed using the NCSS statistical software (NCSS, USA).

3. Results

Basic statistical data for the two contrasting sites including selectedphysico-chemical parameters and bulk concentrations of the studiedcontaminants are given in Table 2. Both sites have similar pH valuesranging from acidic to circumneutral (~4 to 7, mean and median~5)(Table 2). Slightly higher values of Stot, Ccarb and Corg were detectedfor topsoils from the smelting area (Table 2). Similarly, the total

Table 2Basic statistics for selected physico-chemical and chemical parameters of the studied soils.

Code pH (std units) Stot (%) Ccarb

Mining area (Chingola) (n=52) Min 4.16 0.004 0.00Max 7.74 0.336 2.80Mean 5.58 0.047 0.22Q1

a 4.77 0.016 0.04Q2 (Median)a 5.29 0.026 0.07Q3

a 6.46 0.056 0.15Smelting area (Kitwe) (n=55) Min 4.36 0.004 0.03

Max 7.85 0.453 10.4Mean 5.79 0.076 0.88Q1

a 4.93 0.022 0.06Q2 (Median)a 5.59 0.038 0.15Q3

a 6.51 0.074 0.31

a Q1=first quartile=25th percentile; Q2=second quartile=50th percentile (Median);

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

concentrations of metals and As are significantly higher in the smeltingarea (Table 2), indicating the higher exposure of these topsoils relatedto the intense smelter dust fallout. The results of the spatial distributionof As, Co, Cu, Pb and Zn (including selected data fromKříbek et al., 2010)and their bioaccessible fractions are given in Figs. 2 to 6. The bioaccessi-ble fraction of metals and As (expressed as % of the total concentration)is reported in Table 3 and the relationships between the total andbioaccessible concentrations are expressed as correlation coefficientsin Table 4.

Significant differences in As bioaccessibility were observed betweenthe mining and smelting areas, with generally significantly highervalues for the smelter site (Fig. 2, Table 3). The bioaccessible As in themining area varied from 2% to 84% of the total As concentration (medi-an: 9%) and the highest values were observed in the vicinity of theNchanga open pit mine and the Mindolo mine tailing pond (Fig. 2). Incontrast, the As bioaccessibility in the smelting area varied from 19%to 100% of the total As concentration (median: 38%). Topsoils with thehighest As bioaccessibility were located downwind (W and SW) andin the direct vicinity of the smelter, with another hot-spot located inthe low As zone close to the active shafts in the N of the studied area(Fig. 2). Relatively good linear relationships between bioaccessibleand bulk As concentrations were found particularly for the smeltingarea (R=0.973, pb0.001), whereas a lower correlation was found forthe mining area (R=0.878, pb0.001) (Table 4).

The statistical data (Table 3) and Fig. 3 indicate that Co accessibil-ity is similar for both zones. The Co bioaccessibility ranges from 12%to 58% (median: 33%) in the mining area and from 7% to 65% (medi-an: 38%) in the smelting area (Table 3). Zones with the highest Cobioaccessibilities are generally located in the hot-spots correspondingto the highest bulk Co concentrations (R=0.914, pb0.001 for themining area and R=0.931, pb0.001 for the smelting area) (Table 3).

Copper is the most important contaminant with concentrationattaining 10080 ppm in the mining area (median: 457 ppm) and27410 ppm in the smelting area (median: 2027 ppm) (Table 2). Thebioaccessible fraction is similar in both areas, ranging from 38% to83% of the total concentration (median: 58%) and 45–80% (median:60%) in mining and smelting areas, respectively (Table 3). The highestvalues of bioaccessible Cu in the mining area were observed in the vi-cinity of the active mines (N, NE of Chingola city center) and corre-sponded well to topsoils with the highest total Cu concentrations(R=0.973, pb0.001) (Fig. 4 and Table 4). Statistically significant cor-relation was found between total and bioaccessible Cu in the smeltingarea (R=0.997, pb0.001), with the highest bioaccessible fractiondownwind the smelter (Fig. 4 and Table 4).

Lead was found in topsoils in significantly lower concentrationsthan Cu and Co (Table 2), being considered a minor contaminantin the studied areas. Significant differences in Pb bioaccessibilitywere observed between the mining and smelting areas (Fig. 5 andTable 3). Significantly higher Pb bioaccessibilities were found in

(%) Corg (%) As (ppm) Co (ppm) Cu (ppm) Pb (ppm) Zn (ppm)

4 0.09 0.04 2.00 88.0 4.00 6.005.81 5.51 260 10080 63.0 159

5 1.92 1.34 48.3 1380 17.0 33.02 0.94 0.37 7.75 225 4.00 15.05 1.65 0.77 17.5 457 11.0 24.05 2.63 1.39 60.3 1585 22.0 36.50 0.05 0.16 10 365 4.00 7.00

12.8 255 606 27410 480 4503 2.90 9.52 140 4010 35.6 62.70 1.31 1.09 31.5 990 4.00 17.00 2.23 2.91 90 2027 16.0 44.00 3.84 6.26 182 5932 36.5 74.5

Q3=third quartile=75th percentile.

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

Fig. 2. Spatial distribution of As in topsoils, bulk concentrations (ppm) and the bioaccessible fraction.

4 V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

the smelting area, accounting for 25–100% of the total Pb concentra-tion (median: 75%), corresponding well to the zones downwindfrom the smelter stack and with the lowest bioaccessibility valuesupwind and in the vicinity of active mine shafts (Fig. 5). The strongcorrelation between the total and bioaccessible Pb concentrations inthe smelting area was indicated by the high value of the coefficientof correlation (R=0.996, pb0.001) (Table 4). In the mining-affectedarea, the bioaccessible Pb ranged from 11% to 81% of the total Pb

Fig. 3. Spatial distribution of Co in topsoils, bulk conc

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

concentration (median: 40%). Lower statistical relationships wereobserved between bioaccessible and total Pb concentrations (R=0.869,pb0.001), indicating that high values of bioaccessible Pb were alsofound in zones with lower total contents, nevertheless located mostlyin the NE–SW direction, corresponding to the distribution of wind-blown dust from the active mines (Fig. 5).

The bioaccessible fraction of Zn was similar for both studiedareas (Fig. 6). In the mining area, the Zn bioaccessibility ranged

entrations (ppm) and the bioaccessible fraction.

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

Fig. 4. Spatial distribution of Cu in topsoils, bulk concentrations (ppm) and the bioaccessible fraction.

5V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

from 23% to 83% (median: 43) and the relationship between thetotal and bioaccessible Zn was statistically significant (R=0.959,pb0.001) (Tables 3 and 4). In the smelting area the Zn bioaccessibil-ity accounted for 16–79% of the total Zn concentration (median:50%) (Table 3). Slightly lower correlation between the total andbioaccessible Zn concentrations was observed (R=0.946, pb0.001)(Table 4).

Fig. 5. Spatial distribution of Pb in topsoils, bulk conc

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

4. Discussion

4.1. Spatial distribution and bioaccessibility of As and metals in miningand smelting areas

Significant differences in the spatial distribution of metals and Aswere observed between the mining and smelting areas. In addition

entrations (ppm) and the bioaccessible fraction.

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

Fig. 6. Spatial distribution of Zn in topsoils, bulk concentrations (ppm) and the bioaccessible fraction.

6 V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

to old mining activities, emissions from the Nkana smelter at Kitwecontributed to higher concentrations of metals and As (Figs. 2–6,Table 2). The dispersal of contaminants in the vicinity of mines andsmelters is highly dependent on the local meteorological conditions,mainly on the prevailing wind direction (Ettler et al., 2005a, 2011;Kříbek et al., 2010). In addition, dust emitted by the smelters general-ly consists of fine-grained materials with extremely high specificsurface area and high solubility, whereas particles generated by orecrushers in mining areas are generally larger in size (Ettler et al.,2005b, 2008; Kříbek et al., 2010). Thus, the areas affected by smelter-derived particles are generally larger than areas polluted by miningactivities (Figs. 2–6). Kříbek et al. (2010) also emphasized the differ-ences in the chemistry of the dust fallout in mining and smeltingareas. Dust samples collected in the vicinity of open pit mineshave only slightly increased concentrations of Cu (correspondingto traces of chalcopyrite). In contrast, dusts trapped in the vicinityof the Zambian smelters are more enriched in “volatile elements”,such as Pb or As (Kříbek et al., 2010). The relationship between

Table 3Bioaccessible fractions of As and metals expressed as a percentage of total concentra-tions in soils.

Code As Co Cu Pb Zn

Mining area (Chingola) (n=52) Min 2 12 38 11 23Max 84 58 83 81 83Mean 12 34 57 41 45Q1

a 5 28 49 24 34Q2 (Median)a 9 33 58 40 43Q3

a 14 41 64 55 53Smelting area (Kitwe) (n=55) Min 19 7 45 25 16

Max 100 65 80 100 79Mean 40 38 60 73 49Q1

a 28 31 56 67 40Q2 (Median)a 38 38 60 75 50Q3

a 43 49 63 86 57

a Q1=first quartile=25th percentile; Q2=second quartile=50th percentile (Me-dian); Q3=third quartile=75th percentile.

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

these smelter-derived elements is also documented by their statis-tically significant correlation in smelter-affected soils (R=0.896,pb0.001).

However, total concentrations of contaminants are not appropriatefor consideration of metal mobility and bioavailability (Rieuwerts,2007). Thus, metal fractionation studies based on chemical extractionsare often used for this purpose (Ettler et al., 2005a, 2011; Rieuwerts,2007). For example, significantly highermobility of Pb and Zn, expressedas the exchangeable fraction obtained by sequential extraction analysis,was reported by Li and Thornton (2001) at smelting sites in comparisonwith themining sites in theDerbyshire district (England). In soils at othersmelting sites, high percentages of exchangeable (bioavailable) metalsattaining ~50% of the total concentration were also obtained by singleand sequential extractions (Chen et al., 2009; Ettler et al., 2005a, 2011).

Generally, the lower metal and As bioaccessibilities found in themining area close to Chingola (Figs. 2–6; Table 3) are in agreementwith numerous studies dealingwithmining-related soil contamination.It is important to note that the windblown dusts from mine wastes,mine tailing ponds and ore crushers still correspond to themost impor-tant sources of soil pollution in the Zambian mining areas, but are gen-erallymore coarse-grained (N50 μm) than those from smelting facilities(Kříbek et al., 2010; Šráček et al., 2010). Plumlee and Ziegler (2006)state that predominant metal-bearing minerals in mine waste and tail-ings are primarymetal sulphides and sulphosalts and, to a lesser extent,secondary minerals formed by weathering of the ore deposit prior tomining. However, taking into account the hour-scale residence in thestomach, particles containing sulphides should not dissolve substantial-ly under gastric conditions. Similarly, acid-stable Pb sulphates andphosphates should not dissolve to a significant degree (Plumlee andZiegler, 2006; Ruby et al., 1999). The in vivo bioaccessibility studies ofPb uptake by swine also indicated that Pb sulphides and sulphates aresignificantly less dissolved than Pb oxides and carbonates (Casteel etal., 2006; Plumlee et al., 2006).

In contrast, the smelter dusts are generally composed of more solu-ble metal-bearing compounds (Ettler et al., 2005b, 2008). Recent inves-tigations of the pH-dependent leaching behaviour of the copper smelter

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

Table 4Correlation coefficients of the total (M) and bioaccessible As and metal concentrations (Mbio) for the studied soils.

Code Mining area Smelting area

Asbio Cobio Cubio Pbbio Znbio Asbio Cobio Cubio Pbbio Znbio

As 0.878* 0.761* 0.610* 0.309 0.315 0.973* 0.252 0.527* 0.882* 0.643*Co 0.607* 0.923* 0.914* 0.143 0.001 0.483* 0.931* 0.706* 0.395 0.325Cu 0.483* 0.815* 0.973* 0.016 −0.049 0.701* 0.409 0.997* 0.582* 0.375Pb 0.205 0.151 0.008 0.869* 0.522* 0.899* 0.292 0.594* 0.996* 0.870*Zn 0.511* 0.230 0.129 0.528* 0.959* 0.784* 0.383 0.605* 0.957* 0.946*

*Statistically significant correlation at the probability level pb0.001.

Table 5Calculated amounts of contaminant ingested (μg) assuming the soil ingestion rate of100 mg per day for the studied soils.

Code As Co Cu Pb Zn

Mining area (Chingola)(n=52)

Min 0.004 0.20 4.30 0.40 0.30Max 0.08 8.70 579 4.90 10.8Mean 0.01 1.69 84.3 0.896 1.59Median 0.004 0.70 22.3 0.40 0.90

Smelting area (Kitwe)(n=55)

Min 0.01 0.20 19.6 0.40 0.40Max 6.77 36.5 1710 34.2 22.5Mean 0.35 5.90 254 2.69 3.05Median 0.09 3.20 119 1.60 1.90

TDI (μg/day; child 10 kg)a 10 14 1400 36 5000

a TDI=Tolerable daily intake calculated from the human-toxicity maximum permis-sible levels of Baars et al. (2001) in micrograms per day for a child weighting 10 kg.

7V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

flue dust from one smelter in the Zambian Copperbelt indicated thatprimary chalcanthite (CuSO4·5H2O) was readily dissolved, especiallyunder acidic conditions and large amounts of Cuwere released (Vítkováet al., 2011). Similarly, mineralogical analysis of the highly polluted top-soil collected close to the Nkana smelter revealed the presence ofsmelter-derived dust with soluble Cu oxides and sulphates, whichwere responsible for the high vertical mobility in the soil profile (Ettleret al., 2011). Another important parameter of smelter-emitted dust par-ticles is their small grain sizewith diameters generally below10 μmandsubsequent high reactivity in aqueous and soil environments (Ettler etal., 2005b; Vítková et al., 2011). Significantly higher contaminant bioac-cessibilities were also recently reported for the finest fractions of thehighly polluted soils (Juhasz et al., 2011; Madrid et al., 2008a, 2008b).The study by Roussel et al. (2010) was based on investigation of theCd, Pb and Zn bioaccessibilities in soils heavily polluted by Pb-Znsmelters in northern France. They showed that the Pb and Zn gastricbioaccessibilities were between 33% and 76% (median: 65%) and be-tween 17% and 85% (median: 48%), respectively. These data correspondwell to the bioaccessible fractions of Pb and Zn from the studied topsoilsin the smelting area close to Kitwe (Nkana) (Table 3). Although the na-ture of smelter emissions is probably the main reason for higher metal/metalloid bioaccessibilities in the smelting area, the slag particles emit-ted by slag crushers should also not be neglected. Such fine-grainedslags can also be partly dispersed in the vicinity of the Nkana processingcomplex, where old Nkana slags are crushed and transported to theChambishi Co smelter for reprocessing and subsequent Co recovery(Ettler et al., 2011; Kříbek et al., 2010; Vítková et al., 2010). Bosso andEnzweiler (2008) studied the Pb bioaccessibility in highly pollutedsoils from one Brazilian Pb smelting site and found that, under simulat-ed gastrointestinal conditions, an average value corresponding to 70% ofbioaccessible Pb was observed. Morrison and Gulson (2007) investigat-ed the bioaccessibility of metals in base metal smelter slags from NorthLake Macquaire, New South Wales, Australia and found particularlyhigh bioaccessibilities between 80% and 100% for fine grain-size frac-tions (b 20 μm). Together with the small size of particles emitted fromthe smelter stacks and slag reprocessing units (crushers), the bioacces-sibility of some metals/metalloids can be significantly higher in thesmelting areas, as observed in this study (Figs. 2–6; Table 3).

Unfortunately, no soil bioaccessibility data are available in the litera-ture for Co and our study is thefirst investigation of the simulated gastricCo bioaccessibility in the Copperbelt area. Nevertheless, the mineralogi-cal investigations of mining and smelting wastes from the Copperbeltprovince indicated that Co is mainly present as sulphides, intermetalliccompounds and spinels/silicates and to a lesser extent as secondary al-teration products (e.g. carbonates) (Kříbek et al., 2010; Vítková et al.,2010). Thus, compared to other contaminants, a smaller proportion ofCo is mobile (Ettler et al., 2011) and bioaccessible (Fig. 2 and Table 3).This finding is consistent with the fact that Co alloys and spinels werefound to be resistant in the gastric fluids in contrast to Co carbonates,sulphates and oxides (Stopford et al., 2003) (unfortunately, no data areavailable for Co sulphides). Based on this research and previous screen-ing studies (Ettler et al., 2011; Kříbek et al., 2010), themigration and bio-availability of Co in highly polluted soils should be further investigated.

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

4.2. Environmental and health implications

Ruby et al. (1996) in their pioneer study showed that the bioavail-able fraction (i.e. entering the blood stream from the gastrointestinaltract) of Pb and As obtained by a simple in vitro physiologically basedextraction test (PBET) correlated well with in vivo tests. Similarly,Schroder et al. (2004) studied various in vitro methods to predict Pbbioaccessibility in soils and found reasonable agreement with in vivobioavailable Pb estimated from blood data underlining that such sim-ple extraction methods can be used for inexpensive, screening inves-tigation of contaminated soils. Although the simple gastric conditionssimulations (similar to the model used in the present study) arethought to overestimate the total bioaccessibility of metals/metalloidsdue to the aggressive pH of ~1.5 (corresponding to the fasted condi-tions), such bioaccessibility models represent robust tools for humanrisk assessment in areas with high levels of metals/metalloids in soils.Based on the approach of Karadaş and Kara (2011), we calculated thedaily amount of ingested contaminants assuming a soil ingestion rateof 100 mg per day (Table 5). The datawere comparedwith the tolerabledaily intake (TDI) values calculated for a childweighting 10 kgusing thehuman-toxicity maximum permissible levels published by Baars et al.(2001). In particular, Cu and Co in some smelting soil samples exceededand As and Pb approached the TDI values in agreementwith other stud-ies ofmetal bioaccessibility (Juhasz et al., 2011; Karadaş and Kara, 2011;Roussel et al., 2010), indicating again that a higher risk can be expectedin smelter-affected areas. High Co levels in human urine recentlyreported by Banza et al. (2009) in the Cu–Co mining districts of theDemocratic Republic of Congo indicate its relative bioavailability andunderline the importance of ecotoxicological studies in these areas.Our study is the first step in the human risk assessment in the areas ofthemines and smelters of the Zambian Copperbelt and can significantlycontribute to the choice of strategies for reducing human exposure tohigh levels of metals and metalloids in soils. More detailed epidemio-logical studies (similar to those carried out in the nearby miningareas, e.g. Banza et al., 2009) examining the health effects of exposurein various segments of the population, as well as the exact routes of ex-posure (diet, dust ingestion, dust respiration), should be performed.

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001

8 V. Ettler et al. / Journal of Geochemical Exploration xxx (2011) xxx–xxx

5. Conclusions

This study was focused on investigation of the bioaccessibility ofAs and metals (Co, Cu, Pb, Zn) in highly contaminated topsoils fromcontrasting areas in the Zambian Copperbelt (mining- vs. smelting-affected sites). The contaminant bioaccessibility was tested by an invitro method using a simulating gastric fluid containing a 0.4 M solu-tion of glycine adjusted to pH 1.5 by HCl. Significantly higher bioac-cessibilities in the smelter area were observed for As and Pb, attaining100% of the total metal/metalloid concentration. The maximum bioac-cessibilities of As and Pb in the mining area were 84% and 81%, respec-tively. The bioaccessibilities of Co, Cu and Znwere similar for both areas,with maximum values corresponding to 58–65%, 80–83% and 79–83%,respectively. The obtained data and daily intakes calculated for achild weighting 10 kg and assuming a soil intake of 100 mg per dayindicate that a severe health risk related to topsoil ingestion shouldbe taken into account, especially in smelting areas. Direct exposureof inhabitants to high levels of metals (especially Cu and Co) in thesoils of the Zambian Copperbelt must be further evaluated.

Acknowledgements

This study was supported by the Czech Science Foundation (GAČR205/08/0321) and the Ministry of Education, Youth and Sports of theCzech Republic (MSM 0021620855). The research was carried outwithin the framework of IGCP Project No. 594 (“Assessment of impactof mining and mineral processing on the environment and humanhealth in Africa”). Dr. Madeleine Štulíková is thanked for revision ofthe English in the manuscript. Three anonymous reviewers helpedsignificantly to improve the original version of the manuscript.

References

Baars, A.J., Theelen, R.M.C., Janssen, P.J.C.M., Hesse, J.M., van Apeldoorn, M.E., Meijerink,M.C.M., Verdam, L., Zeilmaker, M.J., 2001. Re-evaluation of human-toxicologicalmaximumpermissible risk levels. RIVM report 711701025. Bilthoven, TheNetherlands,p. 297.

Banza, C.L.N., Nawrot, T.S., Haufroid, V., Decrée, S., De Putter, T., Smolders, E., Kabyla, B.I.,Luboya, O.N., Ilunga, A.N., Mtombo, A.M., Nemery, B., 2009. High human exposure tocobalt and other metals in Katanga, a mining area of the Democratic Republic ofCongo. Environmental Research 109, 745–752.

Bosso, S.T., Enzweiler, J., 2008. Bioaccessible lead in soils, slag, and mine wastes from anabandoned mining district in Brazil. Environmental Geochemistry and Health 30,219–229.

Casteel, S.W., Weis, C.P., Henningsen, G.M., Brattin, W.J., 2006. Estimation of relativebioavailability of lead in soil and soil-like materials using young swine. Environ-mental Health Perspectives 114, 1162–1171.

Chen, S., Sun, L., Chao, L., Zhou, Q., Sun, T., 2009. Estimation of lead bioavailability insmelter-contaminated soils by single and sequential extraction procedure. Bulletinof Environmental Contamination and Toxicology 82, 43–47.

Douay, F., Roussel, H., Pruvot, C., Waterlot, C., 2008. Impact of a smelter closedown onmetal contents of wheat cultivated in the neighbourhood. Environmental Scienceand Pollution Research 15, 162–169.

Ettler, V., Vaněk, A., Mihaljevič, M., Bezdička, P., 2005a. Contrasting lead speciation in for-est and tilled soils heavily polluted by lead metallurgy. Chemosphere 58, 1449–1459.

Ettler, V., Johan, Z., Baronnet, A., Jankovský, F., Gilles, C., Mihaljevič, M., Šebek, O.,Strnad, L., Bezdička, P., 2005b. Mineralogy of air-pollution-control residues froma secondary lead smelter: environmental implications. Environmental Scienceand Technology 39, 9309–9316.

Ettler, V., Šebek, O., Grygar, T., Klementová, M., Bezdička, P., Slavíková, H., 2008. Con-trols on metal leaching from secondary Pb smelter air-pollution-control residues.Environmental Science and Technology 42, 7878–7884.

Ettler, V., Johan, Z., Kříbek, B., Šebek, O., Mihaljevič, M., 2009. Mineralogy and environ-mental stability of slags from the Tsumeb smelter, Namibia. Applied Geochemistry24, 1–15.

Ettler, V., Mihaljevič, M., Kříbek, B., Majer, V., Šebek, O., 2011. Tracing the spatial distri-bution and mobility of metal/metalloid contaminants in Oxisols in the vicinity ofthe Nkana copper smelter, Copperbelt province, Zambia. Geoderma 164, 73–84.

International Organization for Standardization, 1995. ISO Standard 11466, Soil quality:extraction of trace elements soluble in aqua regia. International Organization forStandardization, Geneva, Switzerland.

Juhasz, A.L., Weber, J., Smith, E., 2011. Impact of soil particle size and bioaccessibility onchildren and adult lead exposure in peri-urban contaminated soils. Journal of Haz-ardous Materials 186, 1870–1879.

Please cite this article as: Ettler, V., et al., Differences in the bioaccessib(Copperbelt, Zambia), J. Geochem. Explor. (2011), doi:10.1016/j.gexplo

Karadaş, C., Kara, D., 2011. In vitro gastro-intestinal method for the assessment ofheavy metal bioavailability in contaminated soils. Environmental Science and Pol-lution Research 18, 620–628.

Kim, J.-Y., Kim, K.-W., Lee, J.-U., Lee, J.-S., Cook, J., 2002. Assessment of As and heavymetal contamination in the vicinity of Duckum Au-Ag mine, Korean. Environmen-tal Geochemistry and Health 24, 215–227.

Kříbek, B., Majer, V., Veselovský, F., Nyambe, I., 2010. Discrimination of lithogenic andanthropogenic sources of metals and sulphur in soils of the central-northern partof the Zambian Copperbelt Mining District: a topsoil vs. subsurface soil concept.Journal of Geochemical Exploration 104, 69–86.

Li, X., Thornton, I., 2001. Chemical partitioning of trace and major elements in soils con-taminated by mining and smelting activities. Applied Geochemistry 16, 1693–1706.

Madrid, F., Biasoli, M., Ajmone-Marsan, F., 2008a. Availability and bioaccessibility ofmetals in fine particles of some urban soils. Archives of Environmental Contamina-tion and Toxicology 55, 21–32.

Madrid, F., Díaz-Barrientos, E., Madrid, L., 2008b. Availability and bio-accessibility ofmetals in the clay fraction of urban soils of Sevilla. Environmental Pollution 156,605–610.

Mihaljevič, M., Ettler, V., Šebek, O., Sracek, P., Kříbek, B., Kyncl, T., Majer, V., Veselovský,F., 2010. Lead isotopic and metallic pollution record in tree rings from the Copper-belt mining-smelting area, Zambia. Water, Air, and Soil Pollution 216, 657–668.

Morrison, A.L., Gulson, B.L., 2007. Preliminary findings of chemistry and bioaccessibilityin base metal smelter slags. Science of the Total Environment 382, 30–42.

Oomen, A.G., Hack, A., Minekus, M., Zeijdner, E., Cornelis, C., Verstaete, W., Van deWiele, T., Wragg, J., Rompelberg, C.J.M., Sips, A.J.A.M., van Wijnen, J.H., 2002. Com-parison of five in vitro digestion models to study the bioaccessibility of soil con-taminants. Environmental Science and Technology 36, 3326–3334.

Oomen, A.G., Tolls, J., Sips, A.J.A.M., Groten, J.P., 2003a. In vitro intestinal lead uptakeand transport in relation to speciation. Archives of Environmental Contaminationand Toxicology 44, 116–124.

Oomen, A.G., Rompelberg, C.J.M., Bruil, M.A., Dobbe, C.J.G., Pereboom, D.P.K.H., Sips,A.J.A.M., 2003b. Development of an in vitro model for estimating the bioaccessibil-ity of soil contaminants. Archives of Environmental Contamination and Toxicology44, 281–287.

Özkaynak, H., Xue, J., Zartarian, V.G., Glen, G., Smith, L., 2011. Modeled estimates of soiland dust ingestion rates for children. Risk Analysis 31, 592–608.

Pansu, M., Gautheyrou, J., 2006. Handbook of Soil Analysis: Mineralogical, Organic andInorganic Methods. Springer Verlag, Berlin, Heidelberg.

Plumlee, G.S., Ziegler, T.L., 2006. The medical geochemistry of dusts, soils, and otherEarth materials. Treatise in Geochemistry 9, 263–310.

Plumlee, G.S., Morman, S.A., Ziegler, T.L., 2006. The toxicological geochemistry of Earthmaterials: an overview of processes and the interdisciplinary methods used to un-derstand them. Reviews in Mineralogy and Geochemistry 64, 5–57.

Rieuwerts, J.S., 2007. The mobility and bioavailability of trace metals in tropical soils: areview. Chemical Speciation and Bioavailability 19, 75–85.

Roussel, H., Waterlot, C., Pelfrêne, A., Pruvot, C., Mazzuca, M., Douay, F., 2010. Cd, Pband Zn oral bioaccessibility of urban soils contaminated in the past by atmosphericemissions from two lead and zinc smelters. Archives of Environmental Contamina-tion and Toxicology 58, 945–954.

Ruby, M.V., Davis, A., Link, T.E., Schoof, R., Chaney, R.L., Freeman, G.B., Bergstrom, P.,1993. Development of an in vitro screening test to evaluate the in vivo bioaccessi-bility of ingested mine-waste lead. Environmental Science and Technology 27,2870–2877.

Ruby, M.V., Davis, A., Schoof, R., Eberle, S., Sellstone, C.M., 1996. Estimation of lead andarsenic bioavailability using a physiologically based extraction test. EnvironmentalScience and Technology 30, 422–430.

Ruby, M.V., Schoof, R., Brattin, W., Goldade, M., Post, G., Hranois, M., Mosby, D.E., Casteel,S.W., Berti, W., Carpenter, M., Edwards, D., Cragin, D., Chappell, W., 1999. Advancesin evaluating the oral bioavailability of inorganics in soil for use in human healthrisk assessment. Environmental Science and Technology 33, 3697–3705.

Schroder, J.L., Basta, N.T., Casteel, S.W., Evans, T.J., Payton, M.E., Si, J., 2004. Validation ofthe in vitro gastrointestinal (IVG) method to estimate relative bioavailable lead incontaminated soils. Journal of Environmental Quality 33, 513–521.

Soil Survey Staff, 2010. Keys to Soil Taxonomy, 11th Ed. USDA-Natural Resources Con-servation Service, Washington, DC, USA.

Šráček, O., Mihaljevič, M., Kříbek, B., Majer, V., Veselovský, F., 2010. Geochemistryand mineralogy of Cu and Co in mine tailings at the Copperbelt, Zambia. Journalof African Earth Sciences 57, 14–30.

Stopford, W., Turner, J., Cappellini, D., Brock, T., 2003. Bioaccessibility testing of cobaltcompounds. Journal of Environmental Monitoring 5, 675–680.

US EPA, 2007. Estimation of relative bioavailability of lead in soil and soil-like materialsusing in vivo and in vitro methods. Office of Solid Waste and Emergency Response,US EPA, Washington. OSWER 9285.7-77.

VanWijnen, J.H., Clausing, P., Brunekreef, B., 1990. Estimated soil ingestion by children.Environmental Research 51, 147–162.

Vítková, M., Ettler, V., Johan, Z., Kříbek, B., Šebek, O., Mihaljevič, M., 2010. Primary andsecondary phases in copper–cobalt smelting slags from the Copperbelt Province,Zambia. Mineralogical Magazine 74, 581–600.

Vítková, M., Ettler, V., Hyks, J., Astrup, T., Kříbek, B., 2011. Leaching of metals from cop-per smelter flues dust (Mufulira, Zambian Copperbelt). Applied Geochemistry 26,S263–S266.

ility of metals/metalloids in soils from mining and smelting areas.2011.08.001