TECHNICAL REPORTS

2116

In–situ stabilization using phosphate (P) amendments, such as P-based fertilizers and rock, are a potentially cost-eff ective and minimally disruptive alternative for stabilizing Pb in soils. We examined the eff ect of time (0–365 d), in vitro extraction pH (1.5 vs. 2.3), and dosage of three P-based amendments on the bioaccessibility (as a surrogate for oral bioavailability) of Pb in 10 soils from U.S. Department of Defense facilities. Initial untreated soil bioaccessibility consistently exceeded the U.S. Environmental Protection Agency default value of 60% relative bioavailability, with higher bioaccessibility consistently observed at an in vitro extraction pH of 1.5 vs. 2.3. Although P-based amendments statistically (P < 0.05) reduced bioaccessibility in many instances, with reductions dependent on the amendment and dosage, large amendment dosages (approximately 20–25% by mass to yield 5% P by mass) were required to reduce average bioaccessibility by approximately 25%. For most amendment combinations, reductions continued to occur for periods up to 1 yr, indicating that the observed reductions were not merely experimental artifacts of the in vitro extraction procedure. Although our results indicated that reductions in Pb bioaccessibility with P amendments are technically feasible, relatively large amendment masses were required to achieve relatively modest reductions in bioaccessibility. Th e cost and potential environmental implications of adding such large amounts of P may limit the practicality of in situ immobilization for some Pb-contaminated soils, industrial and fi ring range soils in particular.

Decreasing Lead Bioaccessibility in Industrial and Firing Range Soils with Phosphate-

Based Amendments

Rebecca A. Moseley and Mark O. Barnett* Auburn University

Melanie A. Stewart, Tonia L. Mehlhorn, and Philip M. Jardine Oak Ridge National Laboratory

Matthew Ginder-Vogel and Scott Fendorf Stanford University

Lead (Pb) is considered the second most hazardous substance at

contaminated sites in the USA (ATSDR, 2003), having been

identifi ed as a major hazardous chemical at almost one half of the

nation’s 1200+ sites on the National Priority List (Hettiarachchi

and Pierzynski, 2004). Lead poisoning is the most common and

serious environmental health problem facing children in the

USA (Ryan et al., 2004). At Pb-contaminated sites, children can

be exposed to soils through hand-to-mouth activity, where the

ingestion of Pb-contaminated soil is the primary human health

risk exposure pathway (Dudka and Miller, 1999). An alternative

to remediation via the physical excavation and removal of soils

at contaminated sites is to add phosphate (P)-containing acid,

fertilizers, rock, etc. to immobilize Pb (Traina and Laperche,

1999), and in situ stabilization using P-based amendments is

generally regarded as the most cost eff ective and least disruptive

alternative for remediating Pb in soils (Berti and Cunningham,

1997; Ma and Rao, 1999).

Th e primary objective of P-based in situ immobilization is to

reduce the Pb bioavailability (to children on ingestion) to allow-

able levels through the conversion of relatively soluble/bioavailable

forms of Pb to relatively insoluble/less bioavailable pyromorphites

[Pb5(PO

4)

3X(s), where X = Cl−, F−, OH−), the most stable forms

of Pb in oxic surface soils under a wide range of environmental

conditions (Traina and Laperche, 1999). Th e immobilization of

Pb in contaminated soils via P addition has been extensively stud-

ied (Hettiarachchi and Pierzynski, 2004), although many of these

studies have focused on Pb-contaminated urban soils or on Pb-

contaminated soils resulting from mining and smelting.

In addition to urban areas and mining and smelting localities,

other large sources of Pb contamination are industrial activities and

shooting ranges. For example, the U.S. Department of Defense

(DoD) expends more than 2 million pounds of Pb annually at more

than 3000 active small arms fi ring ranges, a number of which are

Abbreviations: BA, amended bioaccessibility; B

U, untreated bioaccessibility, C

A,

amended cleanup level, CU, untreated cleanup level; DI, deionized; DoD, U.S.

Department of Defense; PBET, physiologically based extraction test; RP, rock

phosphate; TIC, total inorganic carbon; TOC, total organic carbon; TSP, triple-super-

phosphate; Vol, VolCanaPhos.

R.A. Moseley and M.O. Barnett, Dep. of Civil Engineering, 208 Harbert Engineering

Ctr., Auburn Univ., Auburn, AL 36849. M.A. Stewart, T.L. Mehlhorn, and P.M. Jardine,

Environmental Sciences Div., P.O. Box 2008, Oak Ridge National Lab., Oak Ridge, TN

37831-6038. M. Ginder-Vogel and S. Fendorf, Dep. of Geological and Environmental

Sciences, Stanford Univ., Stanford, CA 94305.

Copyright © 2008 by the American Society of Agronomy, Crop Science

Society of America, and Soil Science Society of America. All rights

reserved. No part of this periodical may be reproduced or transmitted

in any form or by any means, electronic or mechanical, including pho-

tocopying, recording, or any information storage and retrieval system,

without permission in writing from the publisher.

Published in J. Environ. Qual. 37:2116–2124 (2008).

doi:10.2134/jeq2007.0426

Received 13 Aug. 2008.

*Corresponding author (mark.barnett@auburn.edu).

© ASA, CSSA, SSSA

677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS: HEAVY METALS IN THE ENVIRONMENT

Moseley et al.: Decreasing Pb Bioaccessibility in Soils 2117

facing closure (ITRC, 2003). As a result, Pb is the metal that

most frequently exceeds human health screening criteria at DoD

sites (Salatas et al., 2004). Th ere are an additional approximately

9000 nonmilitary shooting ranges in the USA, and approxi-

mately 4% of total Pb production may go into bullets and shot

(USEPA, 2001). Th ere are similar sources of Pb in other coun-

tries (Johnson et al., 2005; Jorgensen and Willems, 1987; Knech-

tenhofer et al., 2003; Mellor and McCartney, 1994). However,

the eff ect of P on Pb in soils from fi ring ranges and industrial

sites has been studied much less than on soil-bound Pb originat-

ing from urban, mining, and smelting activities.

Th e objectives of this study were to examine the eff ects of

extraction pH, aging time, P source, and P dose on Pb bioac-

cessibility in 10 contaminated soils from DoD facilities. An in

vitro extraction protocol that has been widely used (Berti and

Cunningham, 1997; Brown et al., 2004; Farfel et al., 2005; Het-

tiarachchi and Pierzynski, 2004; Hettiarachchi et al., 2000; Het-

tiarachchi et al., 2001; Hettiarachchi et al., 2003; Sonmez and

Pierzynski, 2005; Yang et al., 2001) to estimate changes in Pb

bioaccessibility induced by soil amendments was used as a surro-

gate for oral bioavailability (Ruby et al., 1996; Ruby et al., 1999).

Methods and Materials

SoilsTen Pb-contaminated soils were collected from various sourc-

es at DoD sites, with Pb concentrations ranging from approxi-

mately 1000 to 6000 mg kg−1 (Table 1). Soils 1 and 2 originated

from a military munitions and small arms testing site, and Soils

7 through 10 came from a fi ring range. Soils 3 through 6 origi-

nated from various defense-related industrial processes: beneath

a fi re tower that had been painted with Pb-based paint (Soil 3),

a settling lagoon (Soil 4), and a sludge pond (Soils 5 and 6). All

soil samples were received air-dry and then sieved to <250 μm,

approximating the particle size ingested by children (Rodriguez et

al., 1999). A particle size of <250 μm was used in determining all

soil properties (with the exception of particle size distribution) as

described by Stewart et al. (2003). Soil pH was determined using

5 mmol L−1 CaCl2 and deionized (DI) water in a 2:1 solution-

to-soil ratio. A microprocessor ionalyzer/901 (Orion Research,

Beverly, MA) with a combination glass and Calomel electrode

(Beckman, Fullerton, CA) was used to measure the pH. Extract-

able iron and manganese oxides were determined with dithionite-

citrate-bicarbonate extraction. Total organic carbon (TOC) and

total inorganic carbon (TIC) were measured by combustion

on a 2400 Series II CHNS/O analyzer (PerkinElmer, Wellesley, MA). Soil TOC was determined on samples pretreated to remove

TIC, which involved a near-boiling 3 mol L−1 HCl extraction.

Soil TIC was computed from the diff erence between total soil C

(no pretreatment) and TOC. Conventional particle size analysis

was used to determine the sand, silt, and clay content of each

soil. Total Pb concentrations in each soil were determined by a

harsh acid digestion Method 3050B (USEPA, 1998) using the

<250-μm particle size fraction.

AmendmentsTh ree common P-rich fertilizers, with a range in solubil-

ity, were used as soil amendments in this study. Triple-super-

phosphate (TSP) and rock phosphate (RP) were obtained from

a local garden center, and VolCanaPhos (Vol) was ordered from

http://www.midwesternbioag.com. Rock phosphate is an in-

soluble, natural rock mined from P-rich sedimentary deposits.

A number of P minerals have been identifi ed in RP, and each

one reacts diff erently because of variations in pH, impurities,

crystalline structure, and local weathering patterns associated

with the location of the deposit (Ma and Rao, 1999). Th e

rock is washed free from clay impurities and heated to remove

moisture. It is then mechanically ground to a fi ne powder to be

incorporated into P-defi cient soils. Rock phosphate naturally

provides long-term, slow-release P (Ma and Rao, 1999). Triple-

super-phosphate, a soluble form of P, is a common agricultural

fertilizer made by reacting RP with orthophosphoric acid (Het-

tiarachchi et al., 2001). Triple-super-phosphate is composed

of 30% monocalcium phosphate (Ca(H2PO

4)

2·H

2O); 45%

gypsum by-product; 10% calcium biphosphate (CaHPO4);

10% iron oxide, silica, and aluminum; and 5% water (Buda-

vari, 1996). VolCanaPhos is the registered trade name for the

crystalline igneous apatite coming from a carbonatite deposit

in northern Ontario, Canada. Carbonatites are carbonate-rich

igneous rocks containing apatite, magnetite, baritite, fl uorite,

Table 1. Properties of study soils.

Soil # Pb Fe Mn TC† TOC TIC‡

pH

Clay Silt Sand 5 mmol L−1 CaCl2

DI§

mg kg−1 g kg−1 mg kg−1 —–––——–%———–––– ——–—–%———––1 4880 11.3 299 1.49 1.23 0.260 6.99 7.47 6.20 30.6 63.2

2 1430 83.9 1160 22.2 36.8 0 6.40 6.79 6.50 14.9 78.6

3 1890 10.2 372 2.59 1.55 1.05 7.52 7.82 11.5 16.5 72.0

4 4660 23.3 87.5 0.640 0.490 0.150 4.11 4.63 19.0 20.0 61.0

5 1060 12.3 458 13.3 15.8 0 7.34 7.70 5.00 20.0 75.0

6 1220 7.90 764 7.68 8.02 0 7.43 7.83 5.00 14.0 81.0

7 1070 19.2 605 1.64 1.44 0.201 6.08 6.52 23.0 33.0 44.0

8 1360 19.3 563 1.83 1.30 0.528 6.14 6.54 13.0 28.0 59.0

9 4020 19.3 490 1.41 1.20 0.204 6.27 6.71 26.0 34.0 40.0

10 5810 17.7 622 1.17 0.984 0.184 6.54 6.95 26.0 35.0 39.0

† TIC, total inorganic carbon; TOC, total organic carbon.

‡ % TIC is obtained by subtracting % TOC from % TC.

§ DI, deionized.

2118 Journal of Environmental Quality • Volume 37 • November–December 2008

P, and irregular concentrations of rare earth elements (USGS,

1995). Like RP, Vol is an insoluble, natural rock that after pelli-

tization provides slow-release P. Th e total P contents (expressed

as weight percent P2O

5) of the RP, TSP, and Vol amendments

were 14.1, 19.8, and 14.1%, respectively.

Amendment Addition and AgingAdding P-based amendments changes the soil’s P content and

potentially the soil pH, both of which can infl uence the bioacces-

sibility of Pb. Th e amendment experiments described below were

designed to examine the overall eff ect of adding P-based amend-

ments to soil, which are generally rototilled or mixed into the

surface of the soil (Brown et al., 2004; Chen et al., 2003; Farfel

et al., 2005; Ryan et al., 2004). Amendment addition and aging

experiments were performed using the <250-μm fraction for both

soil and amendment. Th e air-dried and sieved soil was weighed,

divided into portions, and placed into 20-mL polypropylene

sample vials. Amendments of TSP, RP, and Vol were added to

a fraction of each of the Pb-contaminated soils to produce a

fi nal concentration equivalent to 0% (untreated control), 1, 2.5,

and 5% P by weight (Table 2). Th e normal background total P

concentration in soils varies between 0.02 to 0.10% by weight.

Th erefore, the initial concentration of P in the soils was consid-

ered negligible. Th e fi nal P/Pb ratios were calculated for each soil

for every P dose (Table 3).

Th e P amendments were mixed dry into the soil samples using

a vortex shaker for 1 min. Deionized water was added to bring the

samples to a moisture content of 30%. Th e sample vials were par-

tially covered with perforated parafi lm to allow air exchange while

minimizing evaporation and then placed into a covered container

receiving a continuous fl ow of air at 100% relative humidity. After

the desired aging period, the soils were removed and air dried for

24 h to obtain representative, homogenous samples. Periodic wet-

ting and drying also simulates the variable moisture conditions

experienced by surface soils. However, the eff ect of wetting and

drying on Pb bioaccessibility was not specifi cally examined. After

the samples were completely dried and well mixed, 0.3-g sub-

samples were removed in duplicate for in vitro extraction. Over the

length of the study, several hundred bioaccessibility measurements

were conducted, and the average CV for duplicate samples was

<10%, indicating good sampling, digestion, and analysis repeat-

ability. After subsampling, the remaining samples were re-wet to

30% moisture and placed in the aging apparatus until the next

sampling time. Samples were aged for 365 d, with some intermedi-

ate subsamples taken at 60 and 200 d.

Physiologically Based Extraction TestA streamlined version (Kelley et al., 2002) of the Physi-

ologically Based Extraction Test (PBET), originally developed

by Ruby et al. (1996), was used to measure Pb bioaccessibility.

Th e bioaccessibility of soluble Pb(NO3)

2·6H

2O(s) measured

in the same way was 99.8 ± 1.1% (Yang et al., 2003), so for

this study the absolute and relative bioaccessibility of soil-Pb

were the same. Th e streamlined version is a modifi cation of

the original PBET method in that the NaHCO3 extraction

step (mimicking the small intestine) has been eliminated, re-

fl ecting the results of recent research (Rodriguez et al., 1999;

Ruby et al., 1999). Th e streamlined procedure was originally

designed for 1 g of soil in 100 mL of PBET solution. How-

ever, the procedure was modifi ed to use a <1-g soil sample to

conserve soil mass while maintaining the 1 g:100 mL soil-to-

solution ratio. An experiment was conducted, and replicate

analyses with diff erent sample sizes indicated no signifi cant

diff erence (P < 0.05) in measured Pb bioaccessibility.

Th e primary factor controlling Pb bioaccessibility is the simu-

lated stomach pH (Oomen et al., 2002; Ruby et al., 1996; Ryan

et al., 2004; Yang et al., 2003). Th e PBET extraction was origi-

nally designed to simulate a fasting child’s gastrointestinal tract

at pH 1.5 and a temperature of 37°C (98.6°F). However, Ruby

Table 2. Amendment added expressed as P and total weight percentage.

Amendment† P

Weight of amendment

added

Total weight of soil +

amendmentAmendment

added

% ——————g—————— %

TSP 1 0.101 2.101 4.81

2.5 0.253 2.253 11.2

5 0.505 2.505 20.2

Vol 1 0.142 2.142 6.63

2.5 0.354 2.354 15.0

5 0.709 2.709 26.2

RP 1 0.142 2.142 6.63

2.5 0.355 2.355 15.1

5 0.711 2.711 26.2

† RP, rock phosphate; TSP, triple-super-phosphate; Vol, VolCanaPhos.

Table 3. Lead contents and amended P/Pb ratio.

Soil # P P PbP/Pb ratio (mg/mg)

P/Pb ratio (mol/mol)

% —mg kg−1 soil—1 1 10,000 4880 2.05 13.7

2.5 25,000 4880 5.12 34.3

5 50,000 4880 10.25 68.5

2 1 10,000 1430 6.99 46.8

2.5 25,000 1430 17.5 117

5 50,000 1430 35.0 234

3 1 10,000 1890 5.29 35.4

2.5 25,000 1890 13.2 88.5

5 50,000 1890 26.5 177

4 1 10,000 4660 2.15 14.4

2.5 25,000 4660 5.36 35.9

5 50,000 4660 10.73 71.8

5 1 10,000 1060 9.43 63.1

2.5 25,000 1060 23.6 158

5 50,000 1060 47.2 316

6 1 10,000 1220 8.20 54.8

2.5 25,000 1220 20.5 137

5 50,000 1220 41.0 274

7 1 10,000 1070 9.35 62.5

2.5 25,000 1070 23.4 156

5 50,000 1070 46.7 313

8 1 10,000 1360 7.35 49.2

2.5 25,000 1360 18.4 123

5 50,000 1360 36.8 246

9 1 10,000 4020 2.49 16.6

2.5 25,000 4020 6.22 41.6

5 50,000 4020 12.4 83.2

10 1 10,000 5810 1.72 11.5

2.5 25,000 5810 4.30 28.8

5 50,000 5810 8.61 57.6

Moseley et al.: Decreasing Pb Bioaccessibility in Soils 2119

et al. (1999) noted that pH values of 1.3 and 2.5 in the stomach

phase of the in vitro test correlated equally well with the relative

bioavailability of the in vivo weanling rat model (pH 1.5 was

ultimately selected). Another recent Pb bioavailability study in-

dicated that an in vitro pH of 2.3 correlated well (r2 = 0.90) with

the in vivo rat bone results (Brown et al., 2003). Because bioac-

cessibility is strongly pH dependent, for comparison purposes

bioaccessible Pb was measured at a PBET solution pH of 1.5 at

0, 60, 200, and 365 d and at a PBET pH of 2.3 at 0 and 365 d.

Th e study was thus a full factorial design at all time periods at pH

1.5 and for two time periods for pH 2.3.

Stock PBET solution was made using a 0.4 mol L−1 glycine

(Fisher Scientifi c, Waltham, MA) solution, and the pH was ad-

justed with trace metal grade, 12.1 mol L−1 concentrated hydro-

chloric acid. Th e solution pH was adjusted at a temperature of 37

± 2°C using a pH meter calibrated with buff er solutions adjusted

to a temperature of 37 ± 2°C. Th e extraction test was performed

in a TCLP-like extractor submerged in a heated water bath and

connected to an external motor. Duplicate 0.3-g dry weight sam-

ples were placed into 50-mL polyethylene tubes to which 30 mL

of PBET stock solution heated to 37 ± 2°C at pH 1.5 or 2.3 was

added. Th e pH was adjusted as necessary during the extraction to

maintain the desired pH. Th e samples were rotated end-over-end

at 30 ± 2 rpm for 1 h. During the 1-h extraction period, the wa-

ter temperature in the bath was maintained at body temperature

(37 ± 2°C). After extraction, the samples were removed from the

water bath and centrifuged at 2100 rpm for 5 min to aid with

fi ltration. Th e supernatant was decanted and fi ltered through a

0.45-μm syringe disk fi lter, placed in 20-mL polyethylene vials,

and stored at 4°C until analysis.

Samples containing the dissolved Pb concentration in the

fi ltrate were analyzed in duplicate. A National Institute of

Standards and Technology 2711 standard reference material

was extracted along with the other samples and measured for

Pb with each PBET extraction performed for quality assur-

ance/quality control. Results indicated that conditions in the

PBET were such that the Pb concentration in the standard

reference material was consistently within 10% of the ex-

pected value. Approximately 25% of the remaining PBET

residues were digested by USEPA Method 3050B (USEPA,

1998) and analyzed for Pb to verify a mass balance of ±10%.

Analytical MethodsAll chemicals used in this research were analytical grade

or above, and solutions were prepared with DI water (18

MΩ·cm). Th e PBET supernatant was analyzed for Pb using a

fl ame atomic adsorption spectrometer (SpectrAA 220FS; Var-

ian, Palo Alto, CA) or a PerkinElmer ELAN 6100 (ICP–MS).

Equipment calibration was performed using matrix-matched

standards with a range of 2 to 6 mg Pb L−1. Samples were dilut-

ed gravimetrically with PBET stock solution, if needed, during

analysis to within the concentration range of the standards.

Scanning Electron MicroscopyScanning electron microscopy was used to characterize the mor-

phology and elemental composition of solid-phase Pb in several

samples. After P treatment, dried samples were mounted on con-

ductive carbon tape and were gold coated to limit sample charging.

Back-scattered and secondary electron images were acquired using

a scanning electron microscope (JSM 5600; JEOL, Peabody, MA)

equipped with an energy dispersive detector system.

Results and Discussion

Untreated BioaccessibilityFor children, the USEPA assumes that the absolute bioavailabil-

ity of Pb in diet and water is 50% and that the absolute bioavail-

ability of Pb in soil is 30% (USEPA, 1994). Th is corresponds to a

soil relative absorption factor of 60% for the bioavailability of soil

Pb relative to Pb in water (i.e., relative absorption factor = 0.3/0.5)

(Ruby et al., 1999). Th e bioaccessibility of Pb measured in vitro

depends primarily on the simulated gastric phase, with the simu-

lated gastric pH being the most important variable.

In this study, we tested the eff ect of two gastric pH values,

1.5 and 2.3, on Pb bioaccessibility in these soils (Table 4). We

then compared the EPA default relative bioavailability with the

untreated bioaccessibility measured at pH 1.5 and 2.3 (Fig.

1). Th e average fi nal (365 d) control (without P amendments

added) relative Pb bioaccessibility for all 10 soils at pH 1.5 was

84.8% (range, 55.4–100%). At an in vitro pH of 1.5, 9 of the

10 soils had an initial bioaccessibility greater than the default

value of 60%. Th e untreated Pb bioaccessibility decreased for

all 10 soils as the in vitro pH was increased from 1.5 to 2.3

(average decrease, 24%; range, 8–50%). Despite this decrease,

only 2 of the 10 soils had an initial bioaccessibility less than

the default value of 60% when measured at pH 2.3. Th e aver-

age control relative % Pb bioaccessibility at pH 2.3 was 64.8%

(range, 40.1–85.2%). Using a paired t test, the diff erences

between untreated bioaccessibility measured at pH 1.5 and at

pH 2.3 were signifi cant (P < 0.001), which is consistent with

the results of earlier studies (Oomen et al., 2002; Ruby et al.,

1996; Ryan et al., 2004; Yang et al., 2003). Yang et al. (2002)

demonstrated that the bioaccessibility of As in As-spiked soils

was signifi cantly (P < 0.05) correlated to soil pH and Fe oxide

content. However, in this study, none of the soil properties

(Table 1) was signifi cantly (P > 0.05) correlated with the initial

untreated bioaccessibility at pH 1.5 or 2.3.

Amended BioaccessibilityTo judge the ultimate eff ect of adding P to immobilize Pb in

soil, the relationship between the resulting decrease in bioavail-

ability and its eff ect on site cleanup must be considered. Th e re-

quired health-based cleanup level (C) (e.g., soils with concentra-

tion greater than C must be remediated in some form or fashion)

at a site is inversely proportional to the bioavailability, B.

C α 1/B [1]

Th at is, as the bioavailability goes down, the required cleanup

level increases. Assuming that all of the other parameters in

a given site’s risk assessment (soil ingestion rate, exposure

frequency, body weight, etc.) are constant and can be lumped

into one term, K, Eq. [1] becomes

2120 Journal of Environmental Quality • Volume 37 • November–December 2008

C = K/B [2]

Writing Eq. [2] for an initial (untreated) bioavailability (BU)

with its corresponding required cleanup level (CU) and for a

post-amendment bioavailability (BA) with its corresponding

required cleanup level (CA) and dividing the second by the

fi rst results in

CA/C

U = (K/B

A)/(K/B

U) [3]

or

CA/C

U = 1/(B

A/B

U) [4]

Equation [4] shows that the ratio of the required

post-amendment cleanup level to the required initial,

untreated cleanup level is the inverse of the ratio of the

post-amendment bioavailability to the initial untreated

bioavailability. Equation [4] is nonlinear, which means

that as the amended bioavailability decreases relative to the

untreated bioavailability, the corresponding eff ect on the

required cleanup level increases in a nonlinear fashion (Fig.

2). For example, decreasing the bioavailability by 25% (BA/

BU = 0.75) through soil amendments increases the required

cleanup level by 33% (e.g., from 50 to 67 mg kg−1), whereas

decreasing the bioavailability by 50% (BA/B

U = 0.50) through

soil amendments increases the required cleanup level by a

factor of two (e.g., from 50 to 100 mg kg−1). In this paper we

defi ne the term “marked reduction” to mean relative changes

in bioaccessibility of greater than 25% (i.e., BA/B

U ≤ 0.75).

Amendment addition to contaminated sites involves inherent

uncertainties, costs, and other potential impacts; therefore, it

may be diffi cult to justify remediation via P amendments with

only a relatively modest reduction in bioavailability.

Using this criterion, after 365 d only the highest amendment

dosage (5%) markedly decreased the average bioaccessibility

for these 10 soils measured at an in vitro pH of 1.5 (Table 4).

Although individual soils dosed at 5% P showed reductions in

Table 4. Final (365 d) bioaccessibility of untreated (BU) and amended (B

A/B

U) soil.

TSP† Vol. RP

BU

1% 2.5% 5% 1% 2.5% 5% 1% 2.5% 5%

Soil (%) (BA/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U) (B

A/B

U)

pH 1.5

1 75.5 1.018 0.875 0.679 1.009 0.922 0.812 0.988 0.899 0.830

2 81.5 0.870 0.848 0.775 0.814 0.764 0.662 0.808 0.780 0.673

3 84.4 0.970 0.929 0.792 0.958 0.792 0.680 0.885 0.842 0.808

4 100 0.908 0.847 0.684 0.870 0.776 0.712 0.890 0.831 0.740

5 55.4 0.954 0.927 0.910 0.946 0.857 0.745 0.943 0.911 0.823

6 100 0.878 0.841 0.764 0.849 0.805 0.665 0.825 0.801 0.736

7 95.7 0.774 nd‡ 0.502 0.683 nd 0.578 0.736 nd 0.706

8 94.0 0.768 nd 0.559 0.728 nd 0.442 0.807 nd 0.684

9 75.7 1.071 0.991 0.730 0.855 0.822 0.747 0.926 0.913 0.850

10 85.7 1.052 0.905 0.711 0.989 0.851 0.828 0.979 0.925 0.854

Mean 84.8 0.926 0.895 0.711 0.870 0.824 0.687 0.879 0.863 0.770

Min. 55.4 0.768 0.841 0.502 0.683 0.764 0.442 0.736 0.780 0.673

Max. 100.0 1.071 0.991 0.910 1.009 0.922 0.828 0.988 0.925 0.854

pH 2.3

1 64.22 1.037 0.647 0.272 1.216 1.209 1.184 1.109 0.967 0.891

2 41.01 0.940 0.856 0.716 0.913 0.941 1.013 0.923 0.940 0.803

3 70.71 0.854 0.651 0.427 1.011 1.006 1.003 0.980 0.866 0.643

4 85.18 0.579 0.320 0.208 0.970 0.946 0.954 0.968 0.888 0.821

5 40.14 0.971 0.873 0.635 1.004 1.008 0.936 0.978 0.793 0.523

6 76.63 0.976 0.755 0.492 0.976 1.002 0.913 0.862 0.743 0.523

7 62.92 0.610 nd 0.115 0.814 nd 0.969 0.974 nd 0.624

8 68.24 0.631 nd 0.144 0.851 nd 0.649 0.977 nd 0.632

9 60.65 0.776 0.320 0.178 0.991 1.030 1.097 1.094 1.029 0.853

10 78.76 0.677 0.285 0.175 0.979 0.983 1.058 0.984 0.936 0.819

Mean 64.8 0.805 0.588 0.336 0.973 1.016 0.978 0.985 0.895 0.713

Min. 40.1 0.579 0.285 0.115 0.814 0.941 0.649 0.862 0.743 0.523

Max. 85.2 1.037 0.873 0.716 1.216 1.209 1.184 1.109 1.029 0.891

† RP, rock phosphate; TSP, triple-super-phosphate; Vol, VolCanaPhos.

‡ nd, not determined.

Fig. 1. Untreated/unammended bioaccessibility at pH 1.5 and 2.3. Line indicates USEPA default value of 60% relative bioavailability.

Moseley et al.: Decreasing Pb Bioaccessibility in Soils 2121

bioaccessibility of up to 55%, the average degree of reduction in

Pb bioaccessibility for TSP, Vol, and RP were only 28.9, 31.3,

and 23.0%, respectively. At pH 1.5 and a P dosage of <5%, none

of the soils exhibited a marked reduction in bioaccessibility. Even

when dosed with 5% P, the average Pb bioaccessibility for these

10 soils measured at an in vitro pH of 1.5 was 59.3, 57.6, and

64.8% for TSP, Vol, and RP, respectively, just at the EPA’s default

value of 60%. Adding P to soils can alter bioaccessibilty by react-

ing with the Pb and by changing the soil’s pH. Th us, the results

presented herein refl ect the total eff ect of adding P-based amend-

ments to soils as it is typically practiced in the fi eld.

Th e eff ect of P dose was studied for each amendment on the

365-d data at two pH levels (1.5 and 2.3) by using a paired t test. As with the untreated soils, increasing the PBET pH from

1.5 to 2.3 decreased the Pb bioaccessibility of the amended soils.

Using a paired t test, signifi cant (P < 0.05) decreases in Pb bioac-

cessibility were observed for all P doses for all three amendments

in increasing the PBET pH from 1.5 to 2.3. Th e extraction pH

specifi ed in the PBET procedure is 1.5 (Kelley et al., 2002).

However, pH values of 1.3 and 2.5 correlate equally well with the

relative bioavailability of the in vivo weanling rat model (Ruby et

al., 1999), and recent research has indicated that an in vitro pH

of 2.3 correlates well with in vivo rat bone results (Brown et al.,

2003). In terms of amendment performance, increasing the pH

from 1.5 to 2.3 did not have a consistent eff ect on the eff ective-

ness of the individual amendments. For example, unlike at pH

1.5, there was not a statistically signifi cant (P > 0.05) eff ect of

adding up to 5% P as Vol when measured at pH 2.3. Marked

decreases in Pb bioaccessibility were even observed at 2.5% P

for TSP, the only amendment–pH combination that exhibited

marked reductions in bioaccessibility with <5% P.

One objective of this study was to investigate the relative

bioaccessibility of P-amended, Pb-contaminated soils with re-

spect to time. Th e eff ect of time was examined by comparing

bioaccessibility measurements between two diff erent time points

using a paired t test at pH 1.5 and with a 5% P dose for each

amendment. For all three amendments, there was a signifi cant

diff erence in bioaccessibility measurements over the 365-d period

(Fig. 3–5; Table 5). However, there were signifi cant (P < 0.05)

diff erences in the rate with which the amendments acted.

Th e availability of P is a potential rate-determining step in

the formation of pyromorphite (Ma and Rao, 1999). Triple-

super-phosphate is a soluble P source designed to readily pro-

vide P to soil. Th e P-containing materials in RP are secondary

minerals formed from the alteration of sedimentary materials.

In contrast, the igneous P-containing materials in Vol are pri-

mary minerals. Th e porous nanocrystalline primary minerals

in igneous carbonatites generally have high reactive surface

area compared with the secondary minerals in RP that are

often coated with clays. As such, the solubility and availabil-

ity of P increases in the order RP < Vol < TSP. Th e rate with

which the amendments acted refl ected this order in solubility.

Th e fastest acting amendment was TSP, followed by Vol and

RP. Triple-super-phosphate signifi cantly reduced Pb bioaccessibil-

ity over the time periods 0 to 60 d and 60 to 200 d. However,

there was not a signifi cant diff erence in bioaccessibility measure-

ments for the TSP-amended soils over the fi nal 200 to 365 d,

indicating that the soluble TSP was relatively rapidly acting. In

contrast, insoluble RP was much slower acting. Th ere was not a

signifi cant reduction in measured Pb bioaccessibility until 200 d

for the RP-amended soils, and the bioaccessibility continued to

decrease signifi cantly until the end of the study at 365 d. VolCa-

Fig. 2. Ratio of post-amendment to initial cleanup level (CA/C

U) vs.

post-amendment to initial untreated bioavailability (BA/B

U).

Fig. 3. Lead boaccessibility with respect to time with 5% P as triple-super-phosphate (TSP) at a physiologically based extraction test pH of 1.5. Line indicates USEPA default value of 60% relative bioavailability. The average CV for duplicates over the course of the study was <10%. Error bars are not shown for clarity.

Fig. 4. Lead bioaccessibility with respect to time with 5% P as VolCanaPhos (Vol) at a physiologically based extraction test pH of 1.5. Line indicates USEPA default value of 60% relative bioavailability. The average CV for duplicates over the course of the study was <10%. Error bars are not shown for clarity.

2122 Journal of Environmental Quality • Volume 37 • November–December 2008

naPhos acted at a rate intermediate between TSP and RP, with a

signifi cant decrease occurring between 0 and 60 d, followed by a

plateau in Pb bioaccessibility between 60 and 200 d, and fi nally

by a further signifi cant decrease between 200 and 365 d.

In contrast to our results, several similar studies have report-

ed more limited eff ects of time (Basta et al., 2001; Hettiarach-

chi et al., 2000; Ryan et al., 2004). For example, the work done

by Hettiarachchi et al. (2001) concluded that for soils amended

with RP and TSP, reductions in bioaccessible Pb occurred

between 0 and 3 d after treatment, with no further reductions

occurring over 365 d of incubation. Recently, Scheckel et al.

(2003) concluded that rather than occurring in situ, the desired

Pb-endpoint chloropyromorphite [Pb5(PO

4)

3Cl(s)] can form

as an experimental artifact during sequential extraction proce-

dures, including the PBET (Scheckel et al., 2005). As noted

by Hettiarachchi et al. (2001), the rapid (<3 d) reduction in

bioaccessibility in RP- and TSP-amended soils they observed

may have indeed been attributable to the formation of chlo-

ropyromorphite in the PBET rather than in situ. In contrast,

because Pb bioaccessibility in this study changed signifi cantly

(P < 0.05) over time, the reductions in bioaccessibility cannot

be exclusively due to the formation of chloropyromorphite as

an experimental artifact in the PBET, which is consistent with

results observed by Subacz et al. (2007) for As. We are currently

investigating the formation of Pb5(PO

4)

3Cl(s) in the PBET ex-

traction procedure itself (results not shown).

Solid-Phase Lead Micro-AnalysisSoils #4 and #5 were analyzed using scanning electron micros-

copy. In Soil #4, back-scattered electron images illustrate zones of

Pb that are spatially associated with aluminosilicate clay mineral

surfaces on the basis of EDS analysis (Fig. 6); the intensity of the

BSE image also suggests metallic as well as divalent Pb. Addition-

ally, secondary electron images, coupled with EDS analysis of Soil

#5, reveal Pb associated with aluminosilicate clay minerals (Fig. 7).

Th ese results are consistent with the initial Pb existing as remnant

small metallic fragments and oxidized, cationic Pb that is sorbed

on soil minerals, at least the latter of which was also observed by

Dermatas et al. (2006) in fi ring range soils from DoD sites. Lead-

phosphate precipitates were not observed during the analysis (data

not shown), possibly due to the limited sampling size. Th e appar-

ent predominance of Pb associated with aluminosilicate minerals,

even after treatment with P, suggests that clay minerals may remain

an important reservoir of bioavailable Pb in these soils. Similarly,

chloropyromorphite has evidently not formed in some Pb-contam-

inated soils despite the presence of apatite and favorable geochemi-

cal conditions (Essington et al., 2004).

ConclusionsIn this study, the bioaccessibility of Pb in 10 soils from DoD

facilities was examined as a function of PBET extraction pH,

time, and P amendment dosage. Th e eff ect of extraction pH

was dramatic. At pH 1.5, Pb bioaccessibility was consistently

signifi cantly greater than at pH 2.3. Furthermore, the untreated

Pb bioaccessibility measured at pH 1.5 and pH 2.3 was consis-

tently above the EPA default value of 60%. Th e eff ect of three

P-based amendments on Pb bioaccessibility was also dependent

on time and dosage. As expected, relatively soluble P amend-

ments (TSP) reduced Pb bioaccessibility more rapidly than

relatively nonsoluble amendments (Vol and RP). However,

amendment additions of approximately 20 to 25% by weight

were required to achieve marked (>25%) reductions in Pb bio-

accessibility, and even after amendment, Pb remained bound to

aluminosilicate clay minerals and relatively accessible. Nonethe-

less, the continual reduction in Pb bioaccessibility observed in

these soils over a 1-yr period indicates that at least some of the

reductions in bioaccessibility are occurring in situ rather than

as a result of an experimental artifact, namely the formation of

chloropyromorphite in the PBET extraction. Overall, soluble

TSP was arguably the most eff ective amendment in terms of

time, dosage, and extraction pH. However, these results indi-

cate that relatively large amounts of amendments would need

to be applied in situ to achieve reductions in bioaccessibility of

approximately 25% or more, and similarly modest results have

been reported elsewhere (Hettiarachchi et al., 2000; Hettiarach-

chi et al., 2001; Tang et al., 2004). Th e environmental implica-

tions of adding such large amounts of P to soil, including po-

tential changes in soil structure and enhanced eutrophication of

nearby surface waters (Scheckel and Ryan, 2004) as well as the

mobilization of co-contaminants (Spuller et al., 2007), would

probably also be important. Th us, consistent with the recent

results of Dermatas et al. (2008), adding P amendments may

not be a practical approach to remediate all Pb-contaminated

soils, industrial and fi ring range soils in particular.

Table 5. Eff ect of time (pH 1.5, 5% P) on diff erent P amendments.

Time frame†

Amendment 0–365 d 0–60 d 60–200 d 200–365 d

Vol.‡ yes yes no yes

TSP yes yes yes no

RP yes no no yes

† “Yes” or “no” indicates whether or not there was a signifi cant (P < 0.05)

diff erence in bioaccessibility for the amendment over the timeframe

examined as determined by a paired t test.

‡ RP, rock phosphate; TSP, triple-super-phosphate; Vol, VolCanaPhos.

Fig. 5. Lead bioaccessibility with respect to time with 5% P as rock phosphate (RP) at a physiologically based extraction test pH of 1.5. Line indicates USEPA default value of 60% relative bioavailability. The average CV for duplicates over the course of the study was <10%. Error bars are not shown for clarity.

Moseley et al.: Decreasing Pb Bioaccessibility in Soils 2123

Fig. 6. Backscattered electron image and elemental analysis of Pb-bearing particles in Soil #4.

Fig. 7. Secondary electron images and elemental analysis of Pb-bearing particles in Soil #5.

2124 Journal of Environmental Quality • Volume 37 • November–December 2008

AcknowledgmentsTh is research was sponsored by the Strategic Environmental

Research and Development Program (SERDP) under the direction

of Dr. Andrea Leeson. Th e lead author (RAM) was also supported

in part by a GAANN Fellowship from the U.S. Department of

Education. Th e authors appreciate the assistance of Kent Hartzog

and Vijay Loganathan and the comments of three anonymous

reviewers who greatly improved the manuscript.

ReferencesATSDR. 2003. CERCLA priority list of hazardous substances. Agency for Toxic

Substances and Disease Registry, Atlanta, GA. Available at http://www.atsdr.cdc.gov/cercla/ (verifi ed 11 July 2008).

Basta, N.T., R. Gradwohl, K.L. Snethen, and J.L. Schroder. 2001. Chemical immobilization of lead, zinc, and cadmium in smelter-contaminated soils using biosolids and rock phosphate. J. Environ. Qual. 30:1222–1230.

Berti, W.R., and S.D. Cunningham. 1997. In-place inactivation of Pb in Pb contaminated soils. Environ. Sci. Technol. 31:1359–1364.

Brown, S., R.L. Chaney, J.G. Hallfrisch, and Q. Xue. 2003. Eff ect of biosolids processing on lead bioavailability in an urban soil. J. Environ. Qual. 32:100–108.

Brown, S., R. Chaney, J. Hallfrisch, J.A. Ryan, and W.R. Berti. 2004. In situ soil treatments to reduce the phyto- and bioavailability of lead, zinc, and cadmium. J. Environ. Qual. 33:522–531.

Budavari, S. (ed.) 1996. Merck index. Merck & Co., Whitehouse Station, NJ.

Chen, M., L.Q. Ma, S.P. Singh, R.X. Cao, and R. Melamed. 2003. Field demonstration of in situ immobilization of soil Pb using P amendments. Adv. Environ. Res. 8:93–102.

Dermatas, D., M. Chrysochoou, D.G. Grubb, and X. Xu. 2008. Phosphate treatment of fi ring range soils: Lead fi xation or phosphorous release? J. Environ. Qual. 37:47–56.

Dermatas, D., G. Shen, M. Chrysochoou, D.G. Grubb, N. Menounou, and P. Dutko. 2006. Pb speciation versus TCLP release in army fi ring range soils. J. Hazard. Mater. 136:34–46.

Dudka, S., and W.P. Miller. 1999. Permissible concentrations of arsenic and lead in soils based on risk assessment. Water Air Soil Pollut. 113:127–132.

Essington, M.E., J.E. Foss, and Y. Roh. 2004. Th e soil mineralogy of lead at Horace’s Villa. Soil Sci. Soc. Am. J. 68:979–993.

Farfel, M.R., A.O. Orlova, R.L. Chaney, P.S.J. Lees, C. Rohde, and P.J. Ashley. 2005. Biosolids compost amendment for reducing soil lead hazards: A pilot study of Orgro (R) amendment and grass seeding in urban yards. Sci. Total Environ. 340:81–95.

Hettiarachchi, G.M., and G.M. Pierzynski. 2004. Soil lead bioavailability and in situ remediation of lead-contaminated soils: A review. Environ. Prog. 23:78–93.

Hettiarachchi, G.M., G.M. Pierzynski, F.W. Oehme, O. Sonmez, and J.A. Ryan. 2003. Treatment of contaminated soil with phosphorus and manganese oxide reduces lead absorption by Sprague-Dawley rats. J. Environ. Qual. 32:1335–1345.

Hettiarachchi, G.M., G.M. Pierzynski, and M.D. Ransom. 2000. In situ stabilization of soil lead using phosphorus and manganese oxide. Environ. Sci. Technol. 34:4614–4619.

Hettiarachchi, G.M., G.M. Pierzynski, and M.D. Ransom. 2001. In situ stabilization of soil lead using phosphorus. J. Environ. Qual. 30:1214–1221.

ITRC. 2003. Characterization and remediation of soils at closed small arms fi ring ranges. Interstate Technology and Regulatory Council, Washington, DC.

Johnson, C.A., H. Moench, P. Wersin, P. Kugler, and C. Wenger. 2005. Solubility of antimony and other elements in samples taken from shooting ranges. J. Environ. Qual. 34:248–254.

Jorgensen, S.S., and M. Willems. 1987. Th e fate of lead in soils: Th e transformation of lead pellets in shooting-range soils. Ambio 16:11–15.

Kelley, M.E., S.E. Brauning, R.A. Schoof, and M.V. Ruby. 2002. Assessing oral bioavailability of metals in soil. Battelle Press, Columbus, OH.

Knechtenhofer, L.A., I.O. Xifra, A.C. Scheinost, H. Fluhler, and R. Kretzschmar. 2003. Fate of heavy metals in a strongly acidic shooting-range soil: Small-scale metal distribution and its relation to preferential water fl ow. J. Plant Nutr. Soil Sci. 166:84–92.

Ma, L.Q., and G.N. Rao. 1999. Aqueous Pb reduction in Pb-contaminated soils by

Florida phosphate rocks. Water Air Soil Pollut. 110:1–16.

Mellor, A., and C. McCartney. 1994. Th e eff ects of lead shot deposition on soils and crops at a clay pigeon shooting site in northern England. Soil Use Manage. 10:124–129.

Oomen, A.G., A. Hack, M. Minekus, E. Zeijdner, C. Cornelis, G. Schoeters, W. Verstraete, T. Van de Wiele, J. Wragg, C.J.M. Rompelberg, A. Sips, and J.H. Van Wijnen. 2002. Comparison of fi ve in vitro digestion models to study the bioaccessibility of soil contaminants. Environ. Sci. Technol. 36:3326–3334.

Rodriguez, R.R., N.T. Basta, S.W. Casteel, and L.W. Pace. 1999. An in vitro gastrointestinal method to estimate bioavailable arsenic in contaminated soils and solid media. Environ. Sci. Technol. 33:642–649.

Ruby, M.V., A. Davis, R. Schoof, S. Eberle, and C.M. Sellstone. 1996. Estimation of lead and arsenic bioavailability using a physiologically based extraction test. Environ. Sci. Technol. 30:422–430.

Ruby, M.V., R. Schoof, W. Brattin, M. Goldade, G. Post, M. Harnois, D.E. Mosby, S.W. Casteel, W. Berti, M. Carpenter, D. Edwards, D. Cragin, and W. Chappell. 1999. Advances in evaluating the oral bioavailability of inorganics in soil for use in human health risk assessment. Environ. Sci. Technol. 33:3697–3705.

Ryan, J.A., K.G. Scheckel, W.R. Berti, S.L. Brown, S.W. Casteel, R.L. Chaney, J. Hallfrisch, M. Doolan, P. Grevatt, M. Maddaloni, and D. Mosby. 2004. Reducing children’s risk from lead in soil. Environ. Sci. Technol. 38:18A–24A.

Salatas, J.H., Y.W. Lowney, R.A. Pastorok, R.R. Nelson, and M.V. Ruby. 2004. Metals that drive health-based remedial decisions for soils at U.S. Department of Defense sites. Hum. Ecol. Risk Assess. 10:983–997.

Scheckel, K.G., C.A. Impellitteri, J.A. Ryan, and T. McEvoy. 2003. Assessment of a sequential extraction procedure for perturbed lead-contaminated samples with and without phosphorus amendments. Environ. Sci. Technol. 37:1892–1898.

Scheckel, K.G., and J.A. Ryan. 2004. Spectroscopic speciation and quantifi cation of lead in phosphate-amended soils. J. Environ. Qual. 33:1288–1295.

Scheckel, K.G., J.A. Ryan, D. Allen, and N.V. Lescano. 2005. Determining speciation of Pb in phosphate-amended soils: Method limitations. Sci. Total Environ. 350:261–272.

Sonmez, O., and G.M. Pierzynski. 2005. Phosphorus and manganese oxides eff ects on soil lead bioaccessibility: PBET and TCLP. Water Air Soil Pollut. 166:3–16.

Spuller, C., H. Weigand, and C. Marb. 2007. Trace metal stabilisation in a shooting range soil: Mobility and phytotoxicity. J. Hazard. Mater. 141:378–387.

Stewart, M.A., P.M. Jardine, M.O. Barnett, T.L. Mehlhorn, L.K. Hyder, and L.D. McKay. 2003. Infl uence of soil geochemical and physical properties on the sorption and bioaccessibility of Cr(III). J. Environ. Qual. 32:129–137.

Subacz, J.L., M.O. Barnett, P.M. Jardine, and M.A. Stewart. 2007. Decreasing arsenic bioaccessibility/bioavailability in soils with iron amendments. J. Environ. Sci. Health A 42:1317–1329.

Tang, X.Y., Y.G. Zhu, S.B. Chen, L.L. Tang, and X.P. Chen. 2004. Assessment of the eff ectiveness of diff erent phosphorus fertilizers to remediate Pb-contaminated soil using in vitro test. Environ. Int. 30:531–537.

Traina, S.J., and V. Laperche. 1999. Contaminant bioavailability in soils, sediments, and aquatic environments. Proc. Natl. Acad. Sci. USA 96:3365–3371.

USEPA. 1994. Guidance manual for the integrated exposure uptake biokinetic model for lead in children. U.S. Environmental Protection Agency, Washington, DC.

USEPA. 1998. Test methods for evaluating solid waste, physical/chemical methods. SW-846 SW-846. U.S. Environmental Protection Agency, Washington, DC.

USEPA. 2001. Best management practices for lead at outdoor shooting ranges. EPA-902-B-01-001. U.S. Environmental Protection Agency, Washington, DC.

USGS. 1995. Preliminary compilation of descriptive geoenvironmental mineral deposit models, Open-File Report 95-0831. U.S. Geological Survey, Denver, CO.

Yang, J., D.E. Mosby, S.W. Casteel, and R.W. Blanchar. 2001. Lead immobilization using phosphoric acid in a smelter-contaminated urban soil. Environ. Sci. Technol. 35:3553–3559.

Yang, J.K., M.O. Barnett, P.M. Jardine, and S.C. Brooks. 2003. Factors controlling the bioaccessibility of arsenic(V) and lead(II) in soil. Soil Sed. Contam. 12:165–179.

Yang, J.K., M.O. Barnett, P.M. Jardine, N.T. Basta, and S.W. Casteel. 2002. Adsorption, sequestration, and bioaccessibility of As(V) in soils. Environ. Sci. Technol. 36:4562–4569.