arsenic bioaccessibility in cca-contaminated soils: influence of soil properties, arsenic...

$: Arsenic bioaccessibility in CCA-contaminated soils: Influence of soil properties, arsenic fractionation, and particle-size fraction$
S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 9 ) 2 5 7 6 – 2 5 8 5

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ loca te / sc i to tenv

Arsenic bioaccessibility in CCA-contaminated soils:Influence of soil properties, arsenic fractionation, andparticle-size fraction

Eric Girouard, Gerald J. Zagury⁎

Department of Civil, Geological and Mining Engineering, École Polytechnique de Montréal, Montreal, Québec, Canada H3C 3A7

A R T I C L E D A T A

⁎ Corresponding author. Tel.: +1 514 340 4711xE-mail address: [email protected] (

0048-9697/$ – see front matter © 2008 Elsevidoi:10.1016/j.scitotenv.2008.12.019

A B S T R A C T

Article history:Received 28 March 2008Received in revised form6 December 2008Accepted 10 December 2008Available online 10 February 2009

Arsenic bioaccessibility in soils near chromated copper arsenate (CCA)-treated structureshas recently been reported, and results have shown that soil properties and arsenicfractionation can influence bioaccessibility. Because of the limited data set of publishedresults, additional soil samples and awider range of soil properties are tested in the presentwork. The objectives are: (1) to confirm previous results regarding the influence of soilproperties on arsenic bioaccessibility in CCA-contaminated soils, (2) to investigateadditional soil properties influencing arsenic bioaccessibility, and to identify chemicalextractants which can estimate in vitro gastrointestinal (IVG) bioaccessibility, (3) todetermine arsenic speciation in the intestinal phase of the IVG method and, (4) to assessthe influence of two particle-size fractions on arsenic bioaccessibility. Bioaccessiblearsenic in eight soils collected near CCA-treated utility poles was assessed using the IVGmethod. Five out of the eight soils were selected for a detailed characterization. Moreover,these five soils and two certified reference materials were tested by three different metaloxide extraction methods (citrate dithionite (CD), ammonium oxalate (OX), andhydroxylamine hydrochloride (HH)). Additionally, VMINTEQ was used to determinearsenic speciation in the intestinal phase. Finally, two particle-size fractions (b250 μm,b90 μm) were tested to determine their influence on arsenic bioaccessibility. First, arsenicbioaccessibility in the eight study-soils ranged between 17.0±0.4% and 46.9±1.1% (meanvalue 30.5±3.6%). Using data from 20 CCA-contaminated soil samples, total organic carbon(r=0.50, pb0.05), clay content (r=−0.57, pb0.01), sand content (r=0.48, pb0.05), and water-soluble arsenic (r=0.66, pb0.01) were correlated with arsenic bioaccessibility. The meanpercentage of total arsenic extracted from five selected soils was: HH (71.9±4.1%)NOX(58.0 ±3.1%)Nwater-soluble arsenic (2.2 ± 0.5%), while the mean value for arsenicbioaccessibility was 27.3±2.8% (n=5). Arsenic extracted by HH (r=0.85, pb0.01, n=8) andOX (r=0.93, pb0.05, n=5), showed a strong correlation with arsenic bioaccessibility.Moreover, dissolved arsenic in the intestinal phase was exclusively under the form ofarsenate As(V). Finally, arsenic bioaccessibility (in mg/kg) increased when soil particlesb90 μm were used.

© 2008 Elsevier B.V. All rights reserved.

Keywords:Arsenic speciationOral bioavailabilityBioaccessibilityIn vitroContaminated soilsChromated copper arsenateSoil ingestionParticle-size

4980; fax: +1 514 340 4477.G.J. Zagury).

er B.V. All rights reserved.

mailto:[email protected]

http://dx.doi.org/10.1016/j.scitotenv.2008.12.019

2577S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 9 ) 2 5 7 6 – 2 5 8 5

1. Introduction

Natural soil concentrations of arsenic (As) typically range from0.1 to 40 mg/kg with an average of 5 to 6 mg/kg (WHO, 2001;National Toxicology Program, 2005) and rarely exceed 15mg/kgin North America (Smith et al., 1998). However, high arsenicconcentrations, typically caused by anthropogenic sources, andranging from 10 to N1000 mg/kg have been observed worldwide(Smith et al., 1998).

Seventy percent of the world arsenic production (WHO,2001) and ca. 90% of the USA arsenic production is intended forthe wood preservation industry (decks, playground equip-ment, wood poles, etc.) as chromated-copper-arsenate (CCA)(National Toxicology Program, 2005). CCA is an inorganicwaterborne wood preservative used to extend the service lifeof wood, in which arsenic and copper act as the insecticidesand fungicides respectively. In North America, the mostwidely used formulation of CCA is type C, containing (w/w)47.5% CrO3, 18.5% CuO and 34% As2O5 (Balasoiu et al., 2001).Arsenic in soil and groundwater is a possible threat to humans(Khan et al., 2004), and therefore concerns have been raisedover the potential impact of arsenic leachate in soils onhuman health, particularly children who are most likely tocome in contact with soil (Dagan et al., 2006).

In North America, ingestion of drinking water and food isthe primary route of exposure to arsenic (Belluck et al., 2003).However, incidental ingestion of As contaminated soil is asignificant exposure pathway for children (2 to 6 years old)because of their important hand to mouth activity (Calabreseet al., 1989; Rodriguez et al., 1999; Kwon et al., 2004). In fact,recent studies have shown that incidental ingestion of As-contaminated soil is a major concern (Ljung et al., 2006), andthat exposure to As by dermal absorption and inhalation isconsidered negligible compared to ingestion (Kwon et al., 2004;De Miguel et al., 2007).

Once ingested, the risk for human health is associated withthe fraction of arsenic that is available for absorption intosystemic circulation (Ruby et al., 1993; Rodriguez et al., 1999).Human oral bioavailability is defined as the fraction of thecontaminant that reaches the systemic circulation from thegastrointestinal tract. Bioaccessibility of a contaminant is thefraction that is soluble in the gastrointestinal tract and availablefor absorption. It is, however, unlikely and impractical thatin vivooral bioavailabilitydatawould routinelybegeneratedonasite-specific basis. In vitro methods are currently recognised asrapid screening tools in assessing relative bioavailability ofmetals or metalloids at contaminated sites (Ruby, 2004). Thein vitro gastrointestinal (IVG) method applied on variousnoncalcinated slags and contaminated soils has been success-fully validated for As with in vivo tests using juvenile swine(Rodriguez et al., 1999; Basta et al., 2007). Because of thevariability of soil properties affecting As retention among soils,As bioaccessibility is expected to vary with sites.

To the knowledge of the authors, only one study hasassessed bioaccessibility of As in contaminated soils nearCCA-treated wood poles following incidental soil ingestion.Results suggested that As intake from soil ingestion appearsnegligible compared to the daily intake of inorganic arsenicfrom water and food ingestion for children (Pouschat and

Zagury, 2006). However, this study, performed using a limiteddata set (12 soils), suggested that As bioaccessibility wassystematically higher in coarse-grained soils and in organicsoils. Therefore, additional CCA-contaminated soils must betested to confirm these previous results. Today it is clear that awider range of soil properties (iron (Fe), manganese (Mn) andaluminium (Al) oxides content, phosphorus content, miner-alogy, dissolved organic carbon, etc.) could also influence Asbioaccessibility, as reported in numerous studies (Yang et al.,2002; Ruby, 2004; Cave et al., 2007; Juhasz et al., 2007; Sarkaret al., 2007; Wragg et al., 2007). Hence, a strong need exists forfurther assessment of the influence of soil properties onarsenic bioaccessibility in field-collected CCA-contaminatedsoils, focusing mostly on arsenic association with metaloxides which has been reported in various studies (Manningand Goldberg, 1997; Manning et al., 2002; Rodriguez et al., 2003;Beak et al., 2006; Palumbo-Roe and Klinck, 2007).

Three extraction methods are typically used to quantifyamorphous/crystalline Fe, Mn, Al oxides in soil which canprovide an input to estimate As bioaccessibility: hydroxyla-mine hydrochloride extraction (HH), acid ammonium oxalateextraction (OX) and citrate dithionite extraction (CD). Hydro-xylamine hydrochloride is known to dissolve amorphousmanganese oxide and iron oxide phases, and to releasesurficial adsorbed arsenic and some of the arsenic in themineralmatrix (Rodriguez et al., 2003). The extraction capacityof OX is similar to HH (Chao and Zhou, 1983), and has beenwidely used to dissolve non crystalline forms of iron andaluminum oxides (Carter, 1993). Citrate dithionite extractiondissolves a large proportion of the crystalline iron oxides aswell as much of the amorphous materials and the organiccomplexed iron (Carter, 1993). Citrate dithionite extracted ironoxides (FeCD) were found to be the most important mineralinfluencing adsorption of arsenic in soils (Manning and Gold-berg, 1997), and arsenic bioaccessibility in contaminated soilshas been related to FeCD (Juhasz et al., 2007).

Another important factor to better understand arsenicbioaccessibility is the knowledge of arsenic speciation in thegastrointestinal tract, especially in the intestinal environmentwhere As absorption across the intestinal membrane occurs.The concern is mainly related to the presence of As(III)because of its higher toxicity compared to that of As(V)(Zagury et al., 2008). Thus, a thermodynamic equilibriummodel such as VMINTEQ (Gustafsson, 2006) could be used tomodel the possible As species present in the intestinal extractfrom the IVG method.

Particle-size fraction used in in vitro methods is generallyb250 μm since this fraction adheres more to children's fingersand is thus more available for incidental ingestion (Rodriguezet al., 1999; Yang et al., 2002; Zagury, 2007). A particle-sizefraction b250 μm is also used in the standardized in vitroextraction protocol published by the Solubility BioavailabilityResearch Consortium (Kelley et al., 2002). However, someauthors report that particles adhering to the skin might besmaller than 250 μm (Duggan et al., 1985; Driver et al., 1989;Kissel et al., 1996; Richardson et al., 2006). Furthermore,previous studies on metal/metalloid oral bioavailability eachused a different particle size (b38, b50, b125, b2000 μm): (Hamelet al., 1999; Laird et al., 2007; Ljung et al., 2007; Sarkar et al.,2007;Madrid et al., 2008), while the influence of particle-size on

2578 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 9 ) 2 5 7 6 – 2 5 8 5

arsenic bioaccessibility expressed in mg/kg or as a percentageis not well established (Laird et al., 2007; Ljung et al., 2007).

The objectives of this study are (1) to confirm previousresults pertaining to the influence of soil properties on arsenicbioaccessibility in CCA-contaminated soils using eight sup-plementary soils, (2) to investigate additional soil propertiesinfluencing arsenic bioaccessibility, and to identify chemicalextractants which can estimate IVG bioaccessibility, (3) toprovide input on the geochemical conditions in the intestinalphase of the IVG method and to investigate As speciationthrough geochemical modeling, and (4) to assess the influenceof particle-size fraction on arsenic bioaccessibility.

Additionally, the present study will provide bioaccessibilityand characterization data for two standard reference soils(SRM 2710, CRM 025-050) that will allow comparison withother published in vitro methods, and therefore, will improvethe validation of these protocols.

2. Materials and methods

2.1. Site description, soil samples and certified referencematerials

Eight sites (5, 6, 7, 8, 9, 10, 11 and 12) were selected in theMontreal area (Quebec, Canada) based on soil texture andorganicmatter content. All soil sampleswere collected aroundred pine poles installed between 1988 and 1993, and treatedwith CCA-polyethylene glycol (PEG). In addition, 12 character-ized field-collected CCA-polymer additive (PA) contaminatedsoils from the study of Pouschat and Zagury (2006) wereincluded in the present study.

Soil samples were collected manually with a plastic shovelwithin a 10 cm radius from the pole and at a depth rangingfrom 1 to 10 cm. Samples were placed in 250 ml plasticcontainers and were transported to the laboratory and storedat 4 °C. All instruments used for sample collection weresequentially cleaned with a phosphate-free detergent, rinsedwith deionised water (18.2 MΩ), then with 10% (v/v) nitric acid,and finally rinsed twice with deionised water. All sampleswere dried at 45 °C for 24 h and then sieved in two differentparticle-size fractions (b250 μm, and b90 μm).

To allow comparison with other published results, twocertified reference materials were used (SRM 2710 and CRM025-050). The SRM 2710 (Montana Highly Elevated TraceElement Concentration or Montana I) has already beenextensively tested with various in vitro bioaccessibility meth-ods (Hamel et al., 1999; Oomen et al., 2004; Pouschat andZagury, 2006). The Resource Technology Corporation (RTC)Certified Reference Material CRM 025-050 has been used in aprior study on As bioaccessibility (Pouschat and Zagury, 2006).

2.2. Soil characterization

Soil samples and the certified reference materials werecharacterized for pH, particle-size distribution, total volatilesolids (TVS), total organic carbon (TOC), total phosphorus,total arsenic, total iron, total manganese and gravimetricmoisture content. The pH was measured in distilled wateraccording to ASTM Method D 4972-95a (ASTM, 2004) with a

soil-to-water ratio of 1:2 using a pH electrode (Single-JunctionAg/AgCl reference; Accumet 13-620-285). TVS were deter-mined at 550 °C (Carter, 1993). Particle-size distribution wasperformed according to ASTM Method D422-63 (ASTM, 2004),and the soils were classified using the USDA classificationsystem (gravel [N2mm], sand [2mm–50 µm], silt [50–2 µm] andclay [b2 µm]). The soil gravimetric moisture content wasdetermined by drying at 105 °C according to ASTM Method D2216-98 (ASTM, 2004).

For TOC analysis, samples were ground to powder forhomogenisation and then analyzed by high temperaturecombustion with an induction furnace (LECO, St. Joseph, MI).To determine TOC, the TIC fraction was removed using a 2 MHCl pre-treatment until no effervescence was observed andwas followed by vacuum filtration to removed the CO2

produced (Carter, 1993).Pseudo-total arsenic concentration in the soils was deter-

mined following digestion with a mixture of 8 M HNO3 and2.4 M HCl (CEAEQ, 2006) and analyzed by inductively coupledplasma-atomic emission spectrometry (ICP-AES) (ThermoJarrell Ash, ICAP 61E, Franklin, MA.). Total iron andmanganeseconcentrations in the soils were analyzed by ICP-massspectrometry (ICP-MS) (Perkin Elmer 6100/9000) followingdigestion with HNO3, HClO4 and HF (Tessier et al., 1979).Total phosphorus was measured by colorimetry with theascorbic method (4500-P E.), after digestion according to theMacro-Kjeldahl method (4500-Norg) (Clesceri et al., 1998).

2.3. Chemical extractions

The five selected CCA-contaminated soils (sites 5, 6, 7, 8, and 9)and the two reference soils were extracted by four differentextraction methods. Each method is described below.

2.3.1. Water-soluble arsenic and dissolved organic carbon(DOC) (Rodriguez et al., 2003; Pouschat and Zagury, 2006)For water-soluble arsenic determination, 1 g of soil and 8ml ofdeionised water were placed in a 40 ml polypropylenecopolymer (PPCO) centrifuge tube. The tubes were vigorouslyshaken for 2 h on a Wrist Action Shaker (Burrel Scientific.Pittsburgh, PA; Model 75), and then centrifuged (BeckmanInstruments, Palo Alto, CA; model J2-21) at 7700 ×g for 5 min.Afterwards, the supernatants were filtered (0.45 µm) andanalyzed for As by ICP-MS. The same extraction protocol wasused for DOC determination, analyzing the filtered super-natants with a carbon analyzer (Dohrmann, DC-180).

2.3.2. Hydroxylamine hydrochloride extraction (HH) (Rodriguezet al., 2003)Originally developed by Chao and Zhou (1983), this techniquehas been modified to prevent arsenic readsorption byAmacher and Kotuby-Amacher (1994). One gram of soil(b250 µm) was placed into a 500 ml-PPCO centrifuge bottleand mixed with 250 ml of a solution containing 0.25 M NH2OHHCl, 0.25 MHCl, and 0.025 MH3PO4. The bottles were placed onan orbital shaker (ThermoForma) at 50 °C and shaken for30 min (Pouschat and Zagury, 2006). The samples were thencentrifuged at 6370 ×g for 10 min and the supernatants werefiltered (0.45 µm), stored at 4 °C, and analyzed within 24 h byICP-MS for As, Fe, and Mn.


2.3.3. Acid ammonium oxalate extraction (OX) (McKeagueand Day, 1966; Carter, 1993)One gram of soil (b250 µm) was placed into a 50 ml-PPCOcentrifuge tube and mixed with 50 ml of a solution containing170ml of oxalate solution (0.2 M (NH4)2C2O4·H20) and 107ml ofoxalic acid solution (0.2 M C2H2O4·2H20). The pH was thenadjusted to 3.0 by adding either the oxalate solution or theoxalic acid solution. Subsequently, the tubes were placed on arotary incubator (Environmental Express, model LE1002) for4 h in the dark. The samples were then centrifuged at 4000 ×gfor 15min and the supernatants were filtered (0.45 µm), storedat 4 °C, and analyzed within 24 h by ICP-MS for As, Fe, and Mn.

2.3.4. Citrate dithionite extraction (CD) (Homlgren, 1967;Manning and Goldberg, 1997)One gram of soil (b250 µm) was placed into a 50 ml-PPCOcentrifuge tube and mixed with 50 ml of a solution containing0.57 M Na3C6H5O7·2H2O and 0.1 M Na2S2O4.

The tubeswere placed on a rotary incubator and shaken for16 h. The samples were then centrifuged at 4000 ×g for 15 minand the supernatants were filtered (0.45 µm), stored at 4 °C,and analyzed within 24 h by ICP-MS for As, Fe, and Mn.

2.4. In vitro gastrointestinal (IVG) bioaccessibility

Arsenic bioaccessibility was determined on 13 different soilsamples: 8 CCA-contaminated soils b250 µm, 3 CCA-contami-nated soils b90 µm and 2 certified reference materials, usingthe IVG method (Rodriguez et al., 1999; Pouschat and Zagury,2006). Briefly, 1 g of soil was added to 150ml of gastric solution(containing 0.15MNaCl, Anachemia, Lachine, Qc, Canada, ACSGrade, and 1% w/v pepsin, Sigma-Aldrich, St. Louis, MO, no. P-7000) in a 250 ml beaker placed in a water bath at 37 °C. Thegastric solution pH was adjusted and maintained at 1.80±0.05with environmental-grade HCl throughout the 1-hour gastricphase. During the entire procedure (2 h), argon was constantlybubbled through the solution, and mixing was performedusing individual paddle stirrers (100 rpm). After 1 h, 10 ml ofgastric solution was collected for As analysis. The solutionwas then modified to simulate intestinal solution by adding asaturated NaHCO3 solution to adjust the pH to 5.50±0.05,followed by the addition of porcine bile extract (0.49 g; Sigma-Aldrich, no. B-8631) and porcine pancreatin (0.049 g; Sigma-Aldrich, no. P-1500). After 1 h, 10 ml of intestinal solution wascollected. The 10-ml samples were collected using a new Luer-lock syringe, filtered (0.45 µm), and analyzed for total arsenicby ICP-MS.

Throughout the text, for the sake of simplicity, the termbioaccessibility refers to gastrointestinal bioaccessibilityexpressed in percentage (bioaccessible As/total As in soil).However, because As bioaccessibility (%) is directly influencedby the magnitude of total As content, which can differdepending on particle-size fraction, the gastrointestinalbioaccessibility will also be expressed in terms of bioaccessibleAs concentration (mg/kg) in Section 3.4.

For each experimental method, all labware was sequen-tially cleaned with a phosphate-free detergent, rinsed withdistilled water, then with 10% (v/v) nitric acid, and finallywashed three times with distilled water, and three times withdeionised water.

2.5. Geochemical modeling with VMINTEQ

Additional solution samples (duplicate) were collected andfiltered (0.45 µm) at the end of the intestinal phase for threesoil samples (soil 5, SRM 2710, and CRM 025-050). Thesesampleswere analyzed for DOC, for Na+, Ca2+, K+, Mg2+, As, Mn,Fe, and Al by ICP-MS, and for Cl−, SO4

2−, NO2−, NO3

− by ionicchromatography (IC Dionex DX-100). Alkalinity of the sampleswas also determined by titration with sulphuric acid (Clesceriet al., 1998). During the IVG procedure, redox potential (HachSensIon I) and dissolved oxygen concentration of the solution(Orion Model 1230) were measured every 10 min.

The geochemical model VMINTEQ (Gustafsson, 2006) wasused for speciation analysis. Cations, anions, metals andmetalloids concentrations, pH, temperature, alkalinity andredox potential (Eh) were used as input data. Metal binding tohumic substances was simulated by the NICA-Donnanmodel.It was assumed that 65% of the dissolved organic matter(DOM) was fulvic acid and that 35% was humic acid(Gustafsson and Kleja, 2005). Default hypotheses were usedregarding the nature of the DOC (the ratio of active DOM toDOC was set to 1.4.).

2.6. Quality assurance, quality control

In order to assess the reproducibility of the IVG protocol, twostandard reference materials (SRM 2710 and CRM 025-050)were tested. Both samples had already been tested in aprevious study on As bioaccessibility determined with the IVGprotocol (Pouschat and Zagury, 2006). The gastrointestinalbioaccessibilities were 25.2±0.3% for the SRM 2710 and 64.8±5.2% for the CRM 025-050. The results of the present studywere 23.4±3.0 and 76.0±6.4%, respectively. The results fromboth studies were consistent.

Accuracy and precision of the analytical procedures fortotal As (n=3), total P (n=3), total Fe (n=2), and total Mn (n=2),were verified using the SRM 2710, and the CRM 025-050. TheSRM 2710 has certified values for total As (626±38mg/kg), totalFe (3.38±0.10% w/w), total Mn (1.01±0.04% w/w), and totalP (0.106±0.015% w/w). The CRM 025-050 has certified valuesfor total As (339±51 mg/kg), total Fe (9439±1229 mg/kg), andtotal Mn (173±15 mg/kg). All results obtained in this studywere consistent with the certified values (within 10%).

Thirty-eight procedure blanks were analyzed in parallel forHH (n=6), OX (n=3), CD (n=3), IVG method (n=11), water-soluble As (n=5), DOC (n=5), total Mn and total Fe in soil (n=2),and total As in soil (n=3). All analyzed elements were belowdetection limits in all procedure blanks.

Statistical treatment of the data was performed using theSTATISTICA 7 software (StatSoft, 2004).

3. Results and discussion

3.1. Arsenic bioaccessibility in field collectedCCA-contaminated soils

Arsenic gastrointestinal bioaccessibility in the eight field-collected CCA-PEG contaminated soils (soils 5 to 12) sampledfor the present study ranged between 17.0±0.4% and 46.9±


1.1%,withameanvalueof 30.5±3.6%.Thismeanvaluewasnotsignificantly different than the mean value of 40.7±14.9%obtained from the 12 CCA-PA contaminated soils reported inour previous work (Pouschat and Zagury, 2006) and shown inTable 1. Considering the 20 field collected soils from bothstudies, As gastrointestinal bioaccessibility ranged from 17.0±0.4% to 66.3±2.3%with ameanvalue of 36.6±10.6%. In these 20soils, the highest gastrointestinal bioaccessibilities weremeasured in the three organic soils (Table 2) from sites 2A,2B, 2C (57.3±7.9% on average, n=3) and in the sandy soils fromsites 3A, 3B, 3C and 5 (49.8±3.1% on average, n=4). The lowestbioaccessibilities were measured in a loamy soil with a highclay content fromsite 7 (17.0±0.4%) and a clayey soil fromsite 9(18.2±0.9%).

Also in the same 20 soils, TOC (r=0.50, pb0.05), clay content(r=-0.57, pb0.01), and water-soluble arsenic (r=0.66, pb0.01)were significantly correlated with As bioaccessibility, con-firming previously published results (Pouschat and Zagury,2006). A significant and similar correlation between water-soluble As and arsenic relative bioavailability (RBA) deter-mined in vivo was also observed using mining and smelterwaste samples (r=0.68, pb0.01, n=15) (Rodriguez et al., 2003).Moreover, in the present study, the TVS (r=0.47, pb0.05, n=20)and sand content (r=0.48, pb0.05, n=20) were significantlypositively correlated with arsenic bioaccessibility, while TOC

Table 1 – Fractionation and bioaccessibility of arsenic in 20 soiltwo certified reference materials

Soil Total Asa Water-soluble Asa H

mg/kg %

1Ab 225±6 1.2±0.11Bb 131±0 1.2±0.11Cb 58.0±2.1 1.0±0.12Ab 219±6 15±12Bb 172±19 22±22Cb 153c 35c

3Ab 144±6 2.6±0.13Bb 37.4±2.5 1.3±0.23Cb 231±17 4.1±0.44Ab 251±12 1.4±0.14Bb 173±10 2.1±0.14Cb 238±6 2.1±0.15 141±0 2.2±0.46 134±6 3.6±0.27 310±0 1.6±0.08 178±6 2.4±0.1d

9 125±7 1.4±0.1d

10 193±3 0.9±0.111 103±3 3.8±0.1d

12 195±0 2.3±0.1Mean 170 5.4S.D. 35 2.4Minimum 37.4 0.9Maximum 310 35CRM025-050e 319±12 8.3±0.1

339±51g

SRM 2710f 592±66d 0.06±0.01d

626±38g

aMean value±standard deviation (n=3). bFrom Pouschat and Zagury (2006)values.

and water-soluble As were also correlated (r=0.89, pb0.01,n=20). However, as already reported (Pouschat and Zagury,2006), total arsenic in the soil samples was not correlated witharsenic bioaccessibility. Thus, particle-size distribution,organic matter content and water-soluble As in CCA-con-taminated soils are determining factors that have an effect onarsenic bioaccessibility.

In the eight additional soils sampled for the present study,the arsenic fraction extracted by HH (Table 1) was significantlycorrelatedwith As bioaccessibility (r=0.85, pb0.01, n=8). Theseresults are consistent with the results reported by Rodriguezet al. (2003) who found a significant and similar correlation(r=0.88, pb0.05, n=15) between arsenic extracted by HH and invivo RBA using As-contaminated wastes. However, thesefindings are not in agreement with the debatable resultsreported by Pouschat and Zagury (2006) who extracted nearly100% of arsenic regardless of soil properties.

3.2. Additional soil properties and chemical extractions

3.2.1. Detailed characterization of five study-soil samplesThe five samples (soils 5 to 9) selected for further character-ization (Table 3) had As gastrointestinal bioaccessibilityranging between 17.0±0.4% and 46.9±1.1% with a meanvalue of 27.3±2.8% (Table 1). The five selected soil samples

s (<250 μm) collected near CCA-treated utility poles and in

H extractable Asa Bioaccessible Asa

% % Gastric % Gastrointestinal

98±3 25.7±3.0 28.3±2.296±1 24.2±2.9 26.8±2.899±4 20.7±2.9 25.0±2.793±3 56.2±4.7 59.1±7.079±9 41.7±3.8 46.7±2.894±0 63.6±1.2 66.3±2.3108±10 40.5±3.4 47.7±0.898±9 42.7±2.2 51.2±2.596±9 46.3±0.6 53.5±1.398±5 26.4±1.1 30.9±1.395±6 23.0±1.9 27.1±2.699±3 23.5±1.0 25.9±0.597±2 43.3±3.0 46.9±1.181±2 30.7±3.0 30.9±2.360±2 16.2±1.0 17.0±0.477±3 22.7±1.0 23.5±0.545±0d 15.8±0.0d 18.2±0.9d

82±1 26.4±1.0 31.6±1.491±1 36.2±3.2 45.0±0.761±1 23.6±2.9 30.6±1.787 32.5 36.622 11.2 10.645 15.8 17.0108 63.6 66.397±10 80.4±4.9 76.0±6.4

87±0d 25.7±4.7d 23.4±3.0d

, c(n=1), d(n=2), eCRM025-050 (b850 μm), ffSRM2710 (b74 µm), gCertified

Table 2 – Physicochemical properties of 20 soils (<2 mm) collected near CCA-treated utility poles and of two certifiedreference materials

Soil Texture Sand Silt Clay pHa TVSb TOCa

% % % % (w/w) % (w/w)

1Ac Loam 21.9 25.1 12.2 6.10±0.21b 7.0±0.9 2.61Bc Loam 23.8 32.4 17.8 6.17±0.37b 5.5±1.9 1.51Cc Silt loam 16.8 36.5 17.4 6.37±0.28b 5.9±1.8 2.82Ac Organic 40.4 35.7 1.4 7.13±1.52b 58±30 202Bc Organic 51.2 31.2 6.5 7.17±0.51b 50±30 372Cc Organic 50.0 37.7 4.5 6.20±0.47b 59±44 403Ac Sand 85.9 11.2 2.5 6.18±0.45 3.1±3.0 2.53Bc Sandy loam 47.4 17.8 8.2 6.35±0.43 2.6±2.6 2.33Cc Sand 82.0 11.1 0.9 6.62±0.30 2.0±1.1 2.14Ac Sandy loam 45.8 15.7 10.3 6.64±1.38 4.9±1.8 5.44Bc Loamy sand 62.0 10.8 1.5 5.47±0.49 7.7±4.6 6.14Cc Sandy loam 43.2 30.4 9.5 5.27±0.21 8.0±3.7 5.65 Sand 83.8 4.2 1.0 7.39±0.02 1.9±0.1 0.8±0.06 Sand 81.5 3.0 1.0 7.82±0.01b 1.3±0.0 0.9±0.07 Loam 38.0 30.0 18.0 7.00±0.01b 10.0±0.2 4.2±0.28 Sandy loam 39.8 10.0 4.0 7.75±0.01b 5.1±0.1 2.1±0.29 Clay 18.0 28.0 30.0 7.29±0.01b 9.1±0.0 3.5±0.110 Loam 37.5 27.5 10.0 7.87±0.02 9.5±0.4 4.4±0.111 Loam 47.5 36.2 10.8 7.54±0.02 13±0 5.8±0.112 Organic 43.0 27.0 15.0 6.82±0.02 55±0 24±0Mean – 48.0 23.1 8.7 6.8 16 8.7S.D. – – – – 2.4 62 0.3Minimum – 16.8 3.0 0.9 5.3 1.3 0.8Maximum – 85.9 37.7 30.0 7.9 59 40CRM025-050d Sandy loam – – – 7.34±0.02b 3.7±0.0 0.79±0.12SRM 2710e – – – – 5.06±0.08b 7.8±0.2 2.71±0.04

aMean value±standard deviation (n=3), b(n=2), cFrom Pouschat and Zagury (2006), dCRM 025-050 (b850 μm), eSRM2710 (b74 μm).


represent, according to USDA classification, four different soiltextures: soils 5 and 6 were classified as sand, whereas soils 7,8, and 9 were respectively a loam, a sandy loam, and clay. ThepH values ranged from 7.00±0.01 to 7.82±0.01 (Table 2). Allsoils had a relatively low organic carbon content (TOCb5%)and a low DOC ranging from 56±15 to 214±27 mg/kg (Tables 2and 3). Among the five selected soils, total Fe andMn contentswere lower in the two sandy soils (sites 5 and 6). The low Feand Mn content in these soils is in agreement with the quasiabsence of clay particles (1%) which generally contain Fe andMn oxides. This helps explain the high bioaccessibilitymeasured in these sandy soils.

3.2.2. Chemical extractionsThe mean percent of total arsenic extracted by the chemicalextractants in the five selected CCA-contaminated soils, wasas follows: hydroxylamine hydrochloride (71.9±4.1%)Nammo-nium oxalate (58.0±3.1%)Nwater-soluble arsenic (2.2±0.5%)(Tables 1 and 3). Arsenic extracted by citrate dithionite is notavailable because of analytical problems encountered. It mustbe noted that the two extractants of As bound to Fe,Mn, and Aloxides (HH andOX) overestimated the fraction of bioaccessibleAs (27.3±2.8%). Moreover, in the present study, bioaccessibleAs was significantly correlated with the fraction of arsenicextracted by HH (r=0.89, pb0.05, n=5) and by OX (r=0.93,pb0.05, n=5). As already mentioned, a significant correlation(r=0.88, pb0.05, n=15) between in vivo RBA and As extractedusing HH was also reported (Rodriguez et al., 2003).

Additionally, a significant negative correlation was alsofound between As bioaccessibility and Mn extracted by OX(MnOX) (r=−0.89, pb0.05, n=5) (Table 4). The negative correla-tion between manganese solubilized by OX and arsenicbioaccessibility has not been previously reported. However,some studies have pointed out that Mn can cause a rapidoxidation of As(III) and formation of strongly adsorbed As(V)surface complexes (Manning et al., 2002, 2003).

3.3. Arsenic speciation in the intestinal phase of the IVGmethod

Redox potential (Eh) measured during the IVG method rangedfrom 550–650 mV in the gastric phase (pH=1.8) to 215–325 mVin the intestinal phase (pH=5.5). The dissolved oxygen duringboth phases, ranged from 0.1 to 0.4 mg/l (2–6% saturation) andthe alkalinity of the intestinal solution ranged between 130and 190 mg CaCO3/l.

DOC concentration in the intestinal extracts of the five soilswas extremely high, ranging from 5100 to 6100 mg/l. The lowDOC content (Table 3) of the five selected soils (56–214 mg/kg)confirmed that these high DOC concentrations in the intest-inal extracts originated from the added organic compounds(i.e. pepsin, porcine bile extract and porcine pancreatine). Itwas found that an increase in DOC content promotes both As(V) and As(III) solubilisation (Dobran and Zagury, 2006). Also, asolution containing 25 to 50 mg/l of dissolved organic matter(DOM)mobilized As from the solid phases andmainly targeted

Table 3 – Physicochemical properties of five selected soils (<2 mm) collected near CCA-treated utility poles and of twocertified reference materials

Soil DOCa Total Pa Total Feb Total Mnb OX extractable Asa HH extractable Asa

mg/kg mg/kg mg/kg mg/kg % %

5 108±7 320±100 9200±300 150±1 79±1 97±26 56±15 479±57 9700±100 210±2 70±2 81±27 214±27 1140±40 24,500c 276c 44±1 60±28 107±13b 800±70 24,000±400 438±15 58±2 77±39 200±10b 1470±70 34,900±1400 437±1 37±0b 45±0b

Mean 137 842 20,460 302 58 72S.D. 36 157 1490 15 3 4Minimum 56 320 9200 150 37 45Maximum 214 1470 34,900 438 79 97CRM025-050d 986±28 480±80 8900±220 168±8 91±2 97±10

9439±1229e 173±15e

SRM 2710e 621±0 1180±10 34,000±200 10,100±10 84±1 87±0b

1060±150f 33,800±1000f 10,100±400f

aMean value±standard deviation (n=3), b(n=2), c(n=1), dCRM 025-050 (b850 μm), eSRM2710 (b74 μm), fCertified values.


weakly sorbed arsenic (Bauer and Blodau, 2006). In addition,recent results have shown that natural DOM has a strongpotential to mobilize arsenate from arsenic-loaded ferrihy-drite and Al-ferrihydrite (Mohapatra et al., 2007). Two differenttypes of mechanisms were identified for arsenic release:dissolution of the solid phase and competition betweenarsenic and organic anions for sorption sites (Mohapatraet al., 2007). Therefore, the high DOC concentration measuredin the intestinal extract of the IVG protocol potentially has animportant impact on arsenic bioaccessibility.

Results from geochemical modeling with VMINTEQshowed that the main arsenic species in the intestinal phasewas arsenate: H2AsO4

− (92.3%), HAsO42− (7.6%), H3AsO4 (0.05%).

Dissolved Fe, Al and Mn in solution were mainly bonded withhumic substances. It must be remembered that VMINTEQdatabase does not allow the modeling of binding betweendissolved organic matter and arsenic, and that equilibriumconstants for methylated arsenic species are not included inthe database. It is interesting to note that dissolved metalconcentrationsmeasuredat the endof the IVGmethod intestinal

Table 4 – Extractable elements (mg/kg) of five selected soils (<2certified reference materials

Soil CDa

As Fe Mn As

5 –b 1940±230 28±4 112±16 – 1490±20 70±5 94±27 – 7750±900 195±22 136±38 6700±300 240±8 104±39 – 7670±1450 310±54 48±0.5Mean – 5110 168 99S.D. – 1740 59 5Minimum – 1490 28 48Maximum – 7750 310 136CRM025-050c 301±18 4030±100 139±3 291±5SRM 2710d 557±65 12,700±1300 7030±780 497±3

aMean value±standard deviation (n=3), bNot available cCRM025-050 (b850

phase (data not shown) were very low for Fe (0–1.0 mg/l) and forAl (0.4–0.9 mg/l), whereas dissolved Mn concentrations werehigher (0.2–8.5 mg/l) in accordance with the results obtainedfollowing extraction of Mn oxides. These values also indicatethat Fe, and Al oxides are unlikely to dissolve during the IVGmethod, in agreement with the redox conditions measured.

3.4. Influence of particle-size fraction on as bioaccessibility

Total As content in soils 5, 7 and 8 was influenced by theparticle size fraction but the influence was not consistent(Fig. 1). The sandy soil (site 5) had 3.8 timesmore total As in thesmaller fraction (b90 μm), whereas the loamy soil (site 7)showed a significant reduction in total As in the b90 μm frac-tion (pb0.01). Total arsenic in soil 8 (sandy loam)was similar inparticles b250 μmand in particles b90 μm. It has been recentlyreported that the difference between arsenic content indifferent particle-size fractions was greatest in sandy soilsbecause of the concentration of ions on the limited number ofpreferred binding sites. For soils with a high content of small

50 µm) collected near CCA-treated utility poles and of two

OXa HHa

Fe Mn As Fe Mn

1200±20 31±1 137±3 1760±70 62±2950±50 62±3 108±3 1230±90 136±23500±60 172±13 185±5 5640±30 241±45040±0 156±0 138±5 7590±290 375±243200±40 243±4 56±0 7430±280 443±202780 133 125 4730 25290 14 8 420 32950 31 56 1230 625040 243 185 7590 443595±30 100±3 309±32 1300±70 137±66170±0 5040±140 514±0 8230±3 6170±2

μm), dSRM2710 (b74 μm).

Fig. 2 –Arsenic bioaccessibility (%) in three selectedCCA-contaminated soils for two different particle-sizefractions (<250 µm and <90 µm).


particles, the importance of particle size diminished (Ljunget al., 2006).

For the three selected soils, arsenic gastrointestinal bioac-cessibility (expressed in %) was significantly different betweenthe two particle size fractions (Fig. 2). Arsenic in the sandy soil(site 5) was significantly (pb0.05) less bioaccessible in thesmaller fraction (b90 μm), whereas in this same fractionarsenic was significantly more bioaccessible in the two othersoils. It must be reminded that As bioaccessibility expressedas a percentage (bioaccessible As/total As) is directly influ-enced by themagnitude of total As content. However, for eachof the three samples, the gastrointestinal bioaccessible Asexpressed in mg/kg was significantly higher (pb0.01) in theb90 μm fraction. An in vitro method with three compartments(mouth, stomach, small intestine) applied on two playgroundssoils, also showed higher bioaccessible As (mg/kg) in theb50 µm compared to the b4 mm fraction, but a higher total Ascontent in the b50 µm fraction resulted in a lower percent Asbioaccessibility in this fraction (Ljung et al., 2007). On the otherhand, As bioaccessibility (%) in three mine tailings samplesusing an in vitro method call SHIME, was lower in the smallintestine for the b38 μm fraction compared with the bulkfraction, but the authors noted that the smaller fraction hadgreater total arsenic concentrations (Laird et al., 2007). Thus,with the 3 soil samples tested in this study, a smaller particle-size fraction entailed a higher As bioaccessibility when ex-pressed in mg/kg.

During exposure assessment, arsenic daily intake fromincidental ingestion of soil is calculated using the exposurepoint concentration (soil arsenic concentration) and the RBAwhich can be estimated using in vitro methods. To assessarsenic bioaccessibility, a representative particle-size fractionmust be chosen (generally b250 μm). However, with variousstudies showing the possibility of children ingesting signifi-cant amounts of particles of a smaller size (Duggan et al., 1985;Driver et al., 1989; Kissel et al., 1996; Richardson et al., 2006),the authors suggest that more research is needed on the effectof particle-size distribution on total As content in various soilfractions. Moreover, because smaller soil fractions seem to

Fig. 1 –Total As content and gastrointestinal (GI)bioaccessible arsenic expressed in mg/kg in three selectedCCA-contaminated soils for two different particle-sizefractions (<250 µm and <90 µm).

entail a higher arsenic bioaccessibility, a careful attentionshould be given to the soil fraction to be tested in the assess-ment of bioaccessibility.

4. Conclusions

The following conclusions can be drawn from this study:

• Arsenic bioaccessibility in the eight study-soils rangedbetween 17.0±0.4 and 46.9±1.1% (mean value 30.5±3.6).Considering the soil samples collected near 20 differentCCA-treated utility poles, the average arsenic gastrointest-inal bioaccessibility was 37±11%. These findings clearlyshow that the fraction of arsenic potentially ingested bychildren and solubilized in the gastrointestinal tract ismuchlower than 100%.

• The correlations obtained in the present study confirmedpreviously reported results (Pouschat and Zagury, 2006) andsuggest that organicmatter content (TOC and TVS), particle-size distribution (clay and sand content), and water-solublearsenic are determining factors that significantly influencearsenic bioaccessibility in CCA-contaminated soils.

• None of the tested reagents was able to accurately estimatearsenic bioaccessibility. However, soil arsenic concentrationextracted with hydroxylamine hydrochloride and extractedwith ammonium oxalate were both highly correlated(rN0.89, pb0.05, n=5) with arsenic bioaccessibility, showingthe importance of arsenic bound to amorphous metaloxides. The mean percentage of arsenic extracted fromfive selected CCA-contaminated soils was: hydroxylaminehydrochloride (71.9±4.1%)Nammonium oxalate (58.0±1.3%)Nwater-soluble arsenic (2.2±0.5%), while the meanvalue for arsenic bioaccessibility was 27.3±2.8%. A signifi-cant negative correlation was also found between arsenicbioaccessibility and Mn extracted by ammonium oxalate.

• Geochemical modeling with VMINTEQ suggested that themain arsenic species in the intestinal phase was arsenate:H2AsO4

− (92.3%), HAsO42− (7.6%), H3AsO4 (0.05%).

• Total arsenic content in soils was influenced by the particlesize fractionbut the influencewas inconsistent.Nevertheless,


arsenic bioaccessibility (expressed in mg/kg) was alwaysgreater in the b90 µm than in the b250 µm fraction. It mustbe stressed that the influence of the size fraction on thepercent arsenic bioaccessibility is directly influenced by thetotal arsenic concentration. Therefore, the authors suggest, inagreement with recently published work (Laird et al., 2007),that researchers report both the arsenic concentration andbioaccessible arsenic fraction among various size fractions(b250 µm and smaller if appropriate) for use in human healthrisk assessment depending on the exposure scenario.

Acknowledgments

The authors gratefully acknowledge the financial supportfrom Bell Canada, Hydro Quebec and the Natural Sciences andEngineering Research Council of Canada (NSERC). Thanks arealso due to Manon Leduc, Lucie Jean, Denis Bouchard, RobinPotvin, Mathieu Villeneuve, Dr. Raphaël Mermillod-Blondin,and Dr. John W. Molson.

R E F E R E N C E S

Amacher MC, Kotuby-Amacher J. Selective extraction of arsenicfrom mine spoils, soils and sediments. In: ASA, editor. 1994agronomy abstracts. Madison, WI: ASA; 1994. p. 256.

ASTM. D4972-95a; D422-63; D2216-98 — Annual Book of ASTMStandards. 04.08; 2004.

Balasoiu CF, Zagury GJ, Deschenes L. Partitioning and speciation ofchromium, copper, and arsenic in CCA-contaminated soils:influence of soil composition. Sci Total Environ2001;280:239–55.

Basta NT, Foster JN, Dayton EA, Rodriguez RR, Casteel SW. Theeffect of dosing vehicle on arsenic bioaccessibility insmelter-contaminated soils. J Environ Sci Health Part A2007;42:1275–81.

Bauer M, Blodau C. Mobilization of arsenic by dissolved organicmatter from iron oxides, soils and sediments. Sci Total Environ2006;354:179–90.

Beak DG, Basta NT, Scheckel KG, Traina SJ. Bioaccessibility ofarsenic(V) bound to ferrihydrite using a simulatedgastrointestinal system. Environ Sci Technol 2006;40:1364–70.

Belluck DA, Benjamin SL, Baveye P, Sampson J, Johnson B.Widespread arsenic contamination of soils in residential areasand public spaces: an emerging regulatory or medical crisis?Int J Toxicol 2003;22:109–28.

Calabrese EJ, Barnes R, Stanek EJ, Pastides H, Gilbert CE, VenemanP, et al. How much soil do young-children ingest — anepidemiologic-study. Regul Toxicol Pharmacol 1989;10:123–37.

Carter MR. Soil sampling andmethods of analysis. Boca Raton, Fla:Lewis Publishers; 1993.

Cave M, Taylor H, Wragg J. Estimation of the bioaccessible arsenicfraction in soils using near infrared spectroscopy. J Environ SciHealth Part A 2007;42:1293–301.

CEAEQ. Détermination desmétaux: méthode par spectrométrie demasse à source ionisante au plasma d'argon. MA.200-Mét. 1.1.Québec, Qc, Canada: Ministère de l'Environnement du Québec;2006.

Chao TT, Zhou L. Extraction techniques for selective dissolution ofamorphous iron oxides from soils sediments. Soil Sci Soc Am J1983:225–32.

Clesceri LS, Greenberg AE, Eaton AD. Standard methods for theexamination of water and wastewater. Washington, D.C.:American Public Health Association; 1998.

Dagan R, Bitton G, Townsend TG. Metal transport andbioavailability in soil contaminated with CCA-treated woodleachates. Soil Sediment Contam 2006;15:61–72.

De Miguel E, Iribarren I, Chacon E, Ordonez A, Charlesworth S.Risk-based evaluation of the exposure of children to traceelements in playgrounds in Madrid (Spain). Chemosphere2007;66:505–13.

Dobran S, Zagury GJ. Arsenic speciation and mobilization inCCA-contaminated soils: influence of organic matter content.Sci Total Environ 2006;364:239–50.

Driver JH, Konz JJ, Whitmyre GK. Soil adherence to human-skin.Bull Environ Contam Toxicol 1989;43:814–20.

Duggan MJ, Inskip MJ, Rundle SA, Moorcroft JS. Lead in playgrounddust and on the hands of schoolchildren. Sci Total Environ1985;44:65–79.

Gustafsson JP. Visual MINTEQ. Department of Land and WaterResources Engineering, Stockholm; 2006.

Gustafsson JP, Kleja DB. Modeling salt-dependent protonbinding by organic soils with the NICA—Donnan andStockholm Humic models. Environ Sci Technol2005;39:5372–7.

Hamel SC, Ellickson KM, Lioy PJ. The estimation of thebioaccessibility of heavy metals in soils using artificialbiofluids by two novel methods: mass-balance and soilrecapture. Sci Total Environ 1999;244:273–83.

Homlgren GGS. A rapid citrate–dithionite extractable ironprocedure. Soil Sci Am Procedure 1967;31:210–1.

Juhasz AL, Smith E, Weber J, Rees M, Rofe A, Kuchel T, et al. In vitroassessment of arsenic bioaccessibility in contaminated(anthropogenic and geogenic) soils. Chemosphere2007;69:69–78.

Kelley ME, Brauning SE, Schoof RA, Ruby MV. Assessing oralbioavailability of metals in soil. Columbus, Ohio, USA: BatellePress; 2002.

Khan BI, Solo-Gabriele HM, Dubey BK, Townsend TG, Cai Y.Arsenic speciation of solvent-extracted leachate from new andweathered CCA-treated wood. Environ Sci Technol2004;38:4527–34.

Kissel JC, Richter KY, Fenske RA. Factors affecting soil adherenceto skin in hand-press trials. Bull Environ Contam Toxicol1996;56:722–8.

Kwon E, Zhang HQ, Wang ZW, Jhangri GS, Lu XF, Fok N, et al.Arsenic on the hands of children after playing in playgrounds.Environ Health Perspect 2004;112:1375–80.

Laird BD, VandeWiele TR, Corriveau MC, Jamieson HE, ParsonsMB, Verstraete W, et al. Gastrointestinal microbes increasearsenic bioaccessibility of ingested mine tailings using thesimulator of the human intestinal microbial ecosystem.Environ Sci Technol 2007;41:5542–7.

Ljung K, Selinus O, Otabbong E, Berglund M. Metal and arsenicdistribution in soil particle sizes relevant to soil ingestion bychildren. Appl Geochem 2006;21:1613–24.

Ljung K, Oomen A, Duits M, Selinus O, BerglundM. Bioaccessibilityof metals in urban playground soils. J Environ Sci Health Part A2007;42:1241–50.

Madrid F, Biasioli M, Ajmone-Marsan F. Availability andbioaccessibility of metals in fine particles of some urban soils.Arch Environ Contam Toxicol 2008;55:21–32.

Manning BA, Goldberg S. Arsenic(III) and arsenic(V) absorption onthree California soils. Soil Science 1997;162:886–95.

Manning BA, Fendorf SE, Bostick B, Suarez DL. Arsenic(III)oxidation and arsenic(V) adsorption reactions on syntheticbirnessite. Environ Sci Technol 2002;36:976–81.

Manning BA, Fendorf SE, Suarez DL. Arsenic(III) complexation andoxidation reactions on soil. Biogeochem of EnvironmentallyImportant Trace Elements, vol. 835. 2003. p. 57–69.

McKeague JA, Day JH. Dithionite and oxalate-extractable Fe and Alas aids in differentiating various classes of soils. Can J Soil Sci1966;46:13–22.


Mohapatra D, Mishra D, Rout M, Chaudhury GR. Adsorptionkinetics of natural dissolved organic matter and its impact onarsenic(V) leachability from arsenic-loaded ferrihydrite andAl-ferrihydrite. J Environ Sci Health A Tox Hazard SubstEnviron Eng 2007;42:81–8.

National Toxicology Program. 11th ROC: arsenic compounds,inorganic. 2007, Research Triangle Park, NC; 2005.

Oomen AG, Rompelberg CJM, Van de Kamp E, Pereboom D, DeZwart LL, Sips A. Effect of bile type on the bioaccessibility of soilcontaminants in an in vitro digestion model. Arch EnvironContamToxicol 2004;46:183–8.

Palumbo-Roe B, Klinck B. Bioaccessibility of arsenic in minewaste-contaminated soils: a case study from an abandonedarsenic mine in SW England (UK). J Environ Sci Health Part A2007;42:1251–61.

Pouschat P, Zagury GJ. In vitro gastrointestinal bioavailability ofarsenic in soils collected near CCA-treated utility poles.Environ Sci Technol 2006;40:4317–23.

Richardson GM, Bright DA, Dodd M. Do current standards ofpractice in Canada measure what is relevant to humanexposure at contaminated sites? II: oral bioaccessibility ofcontaminants in soil. Hum Ecol Risk Assess 2006;12:606–16.

Rodriguez RR, Basta NT, Casteel SW, Pace LW. An in vitrogastrointestinal method to estimate bioavailable arsenic incontaminated soils and solid media. Environ Sci Technol1999;33:642–9.

Rodriguez RR, Basta NT, Casteel SW, Armstrong FP, Ward DC.Chemical extraction methods to assess bioavailable arsenic insoil and solid media. J Environ Qual 2003;32:876–84.

Ruby MV. Bioavailability of soil-borne chemicals: abioticassessment tools. Hum Ecol Risk Assess 2004;10:647–56.

Ruby MV, Davis A, Link TE, Schoof R, Chaney RL, Freman GB, et al.Development of an in vitro screening test to evaluate the in

vivo bioaccessibility of ingested mine waste lead. Environ SciTechnol 1993;27:2870–7.

Sarkar D, Makris KC, Parra-Noonan MT, Datta R. Effect of soilproperties on arsenic fractionation and bioaccessibility incattle and sheep dipping vat sites. Environ Int 2007;33:164–9.

Smith E, Naidu R, Alston AM. Arsenic in the soil environment: areview. Advances in Agronomy, Vol 64. 1998. p. 149–95. 64.

StatSoft. Statistica 7: Data analysis software system, Tulsa, OK,USA; 2004.

Tessier A, Campbell PGC, Bisson M. Sequential extractionprocedure for speciation of particulate trace metals. AnalChem 1979;51:844–51.

WHO. Environmental health criteria 224, arsenic and arseniccompound. Inter-organization programme for the soundmanagement of chemicals, Geneva; 2001.

Wragg J, Cave M, Nathanail P. A Study of the relationship betweenarsenic bioaccessibility and its solid-phase distribution in soilsfrom Wellingborough, UK. J Environ Sci Health Part A2007;42:1303–15.

Yang JK, Barnett MO, Jardine PM, Basta NT, Casteel SW.Adsorption, sequestration, and bioaccessibility of As(V) insoils. Environ Sci Technol 2002;36:4562–9.

Zagury GJ. Comments on “Effect of soil properties on arsenicfractionation and bioaccessibility in cattle and sheep dippingvat sites” by D. Sarkar et al. (2007) Environment International33 (2007) 164–169. Environ Int 2007;33:712–3.

Zagury GJ, Dobran S, Estrela S, Deschenes L. Inorganic arsenicspeciation in soil and groundwater near in-service chromatedcopper arsenate-treated wood poles. Environ Toxicol Chem2008;27:799–807.

arsenic bioaccessibility in cca-contaminated soils: influence of soil properties, arsenic...

Documents