Abiotic reductive dechlorination of chlorinated ethylenes by iron-bearing soil minerals. 1. Pyrite and magnetite

Article abstract of DOI:10.1021/es025836b

Abiotic reductive dechlorination of chlorinated ethylenes (tetrachloroethylene (PCE), trichloroethylene (TCE), cis- dichloroethylene (cis-DCE), and vinyl chloride (VC)) by pyrite and magnetite was characterized in a batch reactor system. Dechlorination kinetics was adequately described by a modified Langmuir-Hinshelwood model that includes the effect of a decreasing reductive capacity of soil mineral. The kinetic rate constant for the reductive dechlorination of target organics at reactive sites of soil minerals was in the range of 0.185 (±0.023) to 1.71 (±0.06) day^-1. The calculated specific reductive capacity of soil minerals for target organics was in the range of 0.33 (±0.02) to 2.26 (±0.06) μM/g and sorption coefficient was in the range of 0.187 (±0.006) to 0.7 (±0.022) mM^-1. Surface area-normalized pseudo-first-order initial rate constants for target organics by pyrite were found to be 23.5 to 40.3 times greater than those by magnetite. Target organics were mainly transformed to acetylene and small amount of chlorinated intermediates, which suggests that β-elimination was the main dechlorination pathway. The dechlorination of VC followed a hydrogenolysis pathway to produce ethylene and ethane. The addition of Fe(II) increased the dechlorination rate of cis-DCE and VC in magnetite suspension by nearly a factor of 10. The results obtained in this research provide basic knowledge to better predict the fate of chlorinated ethylenes and to understand the potential of abiotic processes in natural attenuation.

Full text of DOI:10.1021/es025836b

Environ. Sci. Technol. 2002, 36, 5147-5154
lism (15, 16). It has been shown that high contaminant
Abiotic Reductive Dechlorination of
Chlorinated Ethylenes by
concentrations and low temperatures limit the reactivity of
microorganisms (17). Sequential transformation products
(e.g., dichloroethylenes (DCEs) and VC (vinyl chloride))
accumulated by the microbial reductive dechlorination have
been reported to be more toxic and persistent than mother
compounds (3). Abiotic reductive dechlorination has at-
tracted considerable attention in the past decade, due in
part to the limitations of biotic processes.
Iron-Bearing Soil Minerals. 1. Pyrite
and Magnetite
W O O J I N L E E * , † A N D B I L L B A T C H E L O R ‡
Research has been conducted to characterize reductants
that exist in natural environments that can dechlorinate.
Sulfide compounds found in landfill leachate and estuary
sediments have been reported to act as an electron donor
in reductive dechlorination (18-23). NOM (23-25), macro-
cycles (26-29), and mineral surfaces (17, 30-33) have been
shown to significantly increase abiotic transformation rates
of chlorinated organics by acting as electron mediators/
carriers. The reactivity of iron-bearing soil minerals abundant
in natural environments has been investigated as reductants.
Abiotic transformation of carbon tetrachloride (CT), tri-
chloroethane (TCA), trichloroethylene (TCE), and tetra-
chloroethylene (PCE) by iron sulfides has been investigated
using pyrite (32, 34), troilite (35), and mackinawite (36, 37).
The reactivity of sulfide minerals could be caused by Fe(II)
or sulfide on surfaces of the minerals, depending on the
experimental conditions employed (34, 35). It has been shown
that mixed iron oxides such as magnetite can reduce
chlorinated organics. The reaction rate by magnetite has been
reported to be much slower than that by other reactive iron-
bearing soil minerals (35, 38, 39).
Extensive knowledge of the reductive dechlorination
kinetics of target organics and distribution of transformation
products under various environmental conditions is neces-
sary for the successful application and operation of eco-
nomical and effective remedial technologies as well as for
the exact prediction of the fate of chlorinated organics in
natural and engineered systems. The objective of this research
was to systemically characterize the reductive dechlorination
of chlorinated organics by iron-bearing sulfide and oxide
minerals. A series of chlorinated ethylenes (PCE, TCE, cis-
dichloroethylene (cis-DCE), and VC) was chosen as common
environmental contaminants and iron-bearing polysulfide
(pyrite) and oxide (magnetite) were used as representative
soil minerals for the research.
Environmental Science Research Center, School of Public and
Environmental Affairs, Indiana University, Bloomington,
Indiana 47405, and Department of Civil Engineering,
Texas A&M University, College Station, Texas 77843
Abiotic reductive dechlorination of chlorinated ethylenes
(tetrachloroethylene (PCE), trichloroethylene (TCE), cis-
dichloroethylene (cis-DCE), and vinyl chloride (VC)) by pyrite
and magnetite was characterized in a batch reactor
system. Dechlorination kinetics was adequately described
by a modified Langmuir-Hinshelwood model that includes
the effect of a decreasing reductive capacity of soil mineral.
The kinetic rate constant for the reductive dechlorination
of target organics at reactive sites of soil minerals was
in the range of 0.185 ((0.023) to 1.71 ((0.06) day-1. The
calculated specific reductive capacity of soil minerals for
target organics was in the range of 0.33 ((0.02) to 2.26
((0.06) µM/g and sorption coefficient was in the range of
0.187 ((0.006) to 0.7 ((0.022) mM-1. Surface area-
normalized pseudo-first-order initial rate constants for
target organics by pyrite were found to be 23.5 to 40.3 times
greater than those by magnetite. Target organics were
mainly transformed to acetylene and small amount of
chlorinated intermediates, which suggests that â-elimination
was the main dechlorination pathway. The dechlorination
of VC followed a hydrogenolysis pathway to produce
ethylene and ethane. The addition of Fe(II) increased the
dechlorination rate of cis-DCE and VC in magnetite suspension
by nearly a factor of 10. The results obtained in this
research provide basic knowledge to better predict the
fate of chlorinated ethylenes and to understand the potential
of abiotic processes in natural attenuation.
Experimental Section
Anaerobic environments required for this research were
maintained by following the procedures previously described
(40) (see page S1, Supporting Information). Reagents and
samples were prepared in an anaerobic chamber (Coy
Laboratory Products Inc.), and aqueous solutions and
chemicals sensitive to atmospheric oxidation were deoxy-
genated and kept in the anaerobic chamber.
Chem icals. The following chlorinated organics used to
characterize dechlorination kinetics were ACS or higher
grades: PCE (99.9%, Sigma), TCE (99.6%, Sigma), cis-DCE,
(97.0%, Sigma), trans-dichloroethylene (trans-DCE, 98%,
Sigma), 1,1-dichloroethylene (1,1-DCE, 99.0%, Sigma), and
VC (20,000 mg/ L, Sigma). C2 hydrocarbons (ethane (99.0%),
ethylene (1.0%), and acetylene (1,000 mg/ L), Scott Specialty
Gases) were used as standard gases for the measurement of
non-chlorinated transformation products. Target organic
stock solutions were prepared by diluting them in methanol
(99.8%, HPLC grade, EM).
Introduction
Chlorinated ethylenes are widespread groundwater and soil
contaminants in the United States (1-5). Their persistence
in natural environments and toxicity to humans and animals
have been well-known throughout the literature (6-8). Due
to the prevalent occurrence of these contaminants in water
and soil systems and the efforts to treat them, substantial
research has been conducted to identify the mechanisms
that describe their degradation and formation of products
under various environmental conditions. Early research was
mainly focused on the biotic reductive transformation of
chlorinated organics, which generally includes dechlorination
mechanisms such as halorespiration (9-14) and cometabo-
* Corresponding author phone: (812)855-8486; fax: (812)855-1881;
e-mail: woojlee@indiana.edu.
† Indiana University
Deaerated deionized water (ddw) was prepared by
deoxygenating deionized water with 99.99% nitrogen and
‡ Texas A&M University.
9
10.1021/es025836b CCC: $22.00
Published on Web 10/26/2002
2002 Am erican Chem ical Society
VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 5 1 4 7

then by purging with mixed gases (5% hydrogen + 95%
nitrogen) in the anaerobic chamber. Fe(II) stock solution
was prepared by adding an exact amount of FeSO4‚7H2SO4
(102.8%, Sigma) into ddw and kept in the anaerobic chamber
for a maximum of 2 days. A 10 mM buffer solution was
prepared by adding an exact amount of NaHCO3 (100.3%,
Sigma) to ddw. Acid and base solutions were prepared by
diluting H2SO4 (95.7%, Sigma) and NaOH (97.0%, EM) with
ddw.
an aliquot of supernatant was rapidly withdrawn with a
gastight syringe and transferred to an extraction vial with 2
mL of extractant (pentane with toluene as an internal
standard). The target organics and chlorinated transformation
products sorbed on soil minerals was extracted by adding 10
mL of the extractant to the solids that remained after removal
of the supernatant. Extraction was conducted by shaking the
extraction vials on an orbital shaker for 30 min at 250 rpm.
A volume of extractant (1 µL) was injected into split/ splitless
injector at a split ratio of 30:1.
Preparation of Iron-Bearing Soil Minerals. Pyrite (Zacate-
cas, Mexico) was purchased from Ward’s (Rochester, NY)
and ground with ceramic mortar and pestle in the anaerobic
chamber. It was then sieved and pretreated to remove
oxidized soil mineral surfaces. Pyrite was washed with 1 M
acid solution twice and rinsed with ddw several times until
pH of the supernatant was greater than 6.7. It was then freeze-
dried, dry-sieved, and stored in the anaerobic chamber. The
63 to 250 µm size fraction was used for this research.
Magnetite was synthesized by modifying the method de-
veloped by Taylor et al. (41). The synthesized magnetite was
freeze-dried, sieved, and stored under an anaerobic atmo-
sphere. All soil minerals were used for experiments no later
than 7 days after pretreatment and synthesis to preclude an
aging effect (42). The characteristics of pyrite and magnetite
used for this research were reported previously (see pages
S1-S3, Supporting Information).
Non-chlorinated transformation products were analyzed
by a HP 6890 gas chromatograph with a GS-Alumina column
(30 m × 0.53 mm i.d., J&W Scientific) and a flame ionization
detector. The temperature of oven was isothermal at 100 °C.
Supernatant (10 mL) was rapidly transferred with 10-mL
gastight syringe to a 20-mL amber vial. The vial was shaken
for 1 h at 250 rpm and stood for 2 h at room temperature to
equilibrate the aqueous and gas phases. The gas-phase
sample (50 to 100 µL) was withdrawn from the headspace
with a 100 µL gastight syringe (Hamilton) and introduced
into the injector at a split ratio of 5:1. The concentrations of
C2 hydrocarbons in aqueous solution were calculated using
dimensionless Henry’s law constants at room temperature
(20.4, ethane; 8.7, ethylene; 1.1, acetylene) (45, 46).
Chloride was measured by an ion chromatograph (Dionex
500) equipped with AS9-HC column (250 mm × 4 mm i.d.,
Dionex) and conductivity detector. A 10 mM Na2CO3 solution
was used as an eluent and flow rate was 1 mL/ min. A sample
(2.5 mL) of suspension or supernatant was filtered with 0.2
µm membrane filter, diluted if necessary, and injected into
the column through a 10 µL sample loop.
Experim ental Procedures. Batch kinetic experiments
were conducted to characterize dechlorination kinetics and
to identify transformation products. These experiments were
conducted in 20-mL (nominal volume) glass vials with three-
layer seals (PTFE tape, lead foil, PTFE-lined rubber septum)
(43). After equilibration with the anaerobic atmosphere for
2 days, exact amounts of soil minerals were transferred to
the vials followed by a 10 mM NaHCO3 solution (23.6 mL).
Sodium azide (20 mg) was added to one set of soil mineral
suspension to investigate potential microbial activity. A stock
Fe(II) solution (0.5 M FeSO4) was spiked to the magnetite
suspension resulting in Fe(II) concentration of 42.6 mM in
order to investigate the effect of Fe(II) addition on the
dechlorination kinetics of cis-DCE and VC. The magnetite
suspension was exposed to Fe(II) for 1 day in anaerobic
chamber. pH values of soil mineral suspensions were adjusted
to 8 (pyrite) and 7 (magnetite) by adding 1 M acid or base
solutions before the start of experiments and kept constant
during the experiments. The mass ratios of solid to water
were 0.084 (pyrite) and 0.063 (magnetite) resulting in different
surface area concentrations of pyrite (2340 m2/ L) and
magnetite (3600 m2/ L). The average headspace volume of
the vial was 0.6 mL, which would allow less than 1.5% of TCE
partitioning to the headspace assuming a dimensionless
Henry’s law constant of 0.359 at 22 °C (44) and no sorption.
Soil mineral suspensions were spiked with methanolic stock
solutions (10-50 µL) to obtain initial concentrations of 0.19
mM PCE, 0.25 mM TCE, 0.41 mM cis-DCE, and 0.40 mM VC.
The vials were rapidly and tightly capped, taken out of the
anaerobic chamber, mounted on the tumbler, and then
completely mixed at 7 rpm at room temperature (22 ( 0.5
°C). All soil mineral samples and controls (bicarbonate buffer
+target organics) were prepared in duplicate. Target organics
and their transformation products in aqueous solutions and
on soil minerals were measured at each sampling time.
Analytical Procedures. Target organics and their chlo-
rinated transformation products were analyzed by a Hewlett-
Packard (HP) G1800A gas chromatography detector (GCD)
system with a DB-VRX column (60 m × 0.25 mm i.d. × 1.8
µm film thickness, J&W Scientific) and a mass spectrometer
detector (MS/ EID). The oven temperature was isothermal at
80 °C for 8 min, increased to 160 °C at the rate of 20 °C/ min,
and then held for 2 min. To analyze target organics and
chlorinated transformation products in aqueous solution,
Results and Discussion
Treatm ent of Kinetic Data (see pages S4-S5, Supporting
Information). Chlorinated ethylenes in the soil mineral
suspensions showed a rapid disappearance initially, followed
by slower rates of removal or constant concentrations,
depending on the reductive capacity of the soil minerals.
This is different from the behavior of chlorinated organics
in most other abiotic reductive transformations, where the
kinetics are usually described by a pseudo-first-order or zero-
order rate law (22, 32, 34-37, 45, 47). To describe dechlo-
rination kinetics of target organics in the soil mineral
suspensions, it is assumed that 1) target organics adsorb
onto a finite number of reactive sites on the surfaces of soil
minerals, 2) reductive dechlorination of the target organics
occurs at these sites by a first-order reaction resulting in loss
of activity of the sites, and 3) the reactive sites are the source
of reductive capacity of soil minerals for target organic.
Therefore, the decay rate of target organic (rdecay) would be
proportional to the concentration of target organic adsorbed
onto the reactive sites, which could be described by the
Langmuir-Hinshelwood rate law
kCRCCCE
rdecay ) kCtSCE
)
(1)
1/ K + CCE
where k is the rate constant for the decay of target organic
at the reactive sites (day-1); CtSCE is the concentration of
target organic adsorbed on the reactive sites (mM); CCE is the
aqueous concentration of target organic (mM); K is the
sorption coefficient of target organic on reactive sites (mM-1);
and CRC is the concentration of reductive capacity of soil
minerals for target organics (mM). The concentration of
reductive capacity represents the total amount of target
organic that the mineral can possibly reduce per unit volume
of water and is proportional to the concentration of reactive
sites. The reductive capacity of the soil minerals is expressed
in terms of the target organic, because it is assumed that the
9
5 1 4 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002

reactive sites do not react with any transformation products.
Therefore, the concentration of reductive capacity at any
time can be calculated as the difference in the initial reductive
capacity and the change in total concentration of target
organic. The total concentration is obtained by multiplying
the aqueous concentration by partitioning factor
CRC ) CR0C - (CC0 E,Total - CCE,Total) ) CR0C - pCE(CC0 E - CCE
)
(2)
Vg
pCE ) 1 + HCE
+ ks
(3)
(
)
Vaq
where C0RC is the initial concentration of reductive capacity
of soil mineral for the target organic (mM); C0CE,Total is total
concentration of target organic at time equal to zero (mM);
FIGURE 1. Reductive dechlorination of chlorinated ethylenes in
0.084 g/g pyrite suspensions. The error bars are 95% confidence
intervals. Curves represent kinetic model fits. Control is the control
sample for TCE.
C
CE,Total is the total concentration of target organic in all phases
(mM); C0CE is the initial aqueous concentration of target
organic (mM); pCE is the dimensionless equilibrium parti-
tioning factor that equals the ratio of the amount of target
organic in all phases to that in the aqueous phase (43); HCE
is the dimensionless Henry’s law constant for target com-
pounds; Vg and Vaq are the volumes of gas and aqueous
phases; and ks is the dimensionless partition coefficient of
chlorinated ethylenes to the solid phase of reactor system
that are assumed to be nonreactive (see page S6, Supporting
Information). The batch reactor used for this study had
slightly different gas and aqueous phase volumes under
various experimental conditions. The volumes of gas and
aqueous phases were typically 0.6 ( 0.2 mL and 23.7 ( 0.2
mL. CRC in eq 1 can be substituted by the right side of eq 2.
The rate equation can then be described in terms of measured
target organic concentrations as follows.
of initial TCE concentration at 1.5 day and then gradually
decreased over time reaching 94% of the initial value (65.3
days). The half-lives of TCE partitioning to the solid phases
in controls (Teflon septum liner + reactor wall) and that to
solid phases in pyrite suspensions (Teflon septum liner +
reactor wall + pyrite) were less than 0.65 day. The rates of
partitioning of TCE to the solid phases were much faster
than rates of reductive dechlorination of TCE in pyrite
suspension. Therefore, the assumption of instantaneous
equilibrium for target organic partitioning to the solid phase
is reasonable. The concentration of PCE in the control (not
shown) also rapidly dropped and reached equilibrium at 1.5
day (approximately 94% of initial PCE). In cases of cis-DCE
and VC controls (not shown), the concentrations of target
organics were relatively constant (average > 98%), which
indicates that partitioning of cis-DCE and VC to the solid
phases of the reactor system was negligible.
k{C0RC - pCE(C0CE - CCE)}CCE
rdecay
)
(4)
1/ K + CCE
Table 1 shows the kinetic parameters, transformation
products and their recoveries, and target organic remaining
at the last sampling time. Recovery of transformation product
and target organic remaining were calculated by dividing
total concentrations of transformation product and target
organic at the last sampling time by the initial concentration
of target organic. This assumes that the initial target organic
concentration is the maximum possible concentration of
transformation products. The incompleteness of total re-
covery of target organics and transformation products may
be caused by the formation of nondetectable products,
inaccuracy in Henry’s law constants, and volatilization loss
during the sampling procedure.
Pyrite suspensions removed 29.8-50% of initial target
organics, and the carbon mass balance stayed between 88.6
and 99% at the last sampling times. Target organics showed
a pseudo-first-order decay during early reaction time, which
is similar to that observed in the reductive dechlorination of
chlorinated ethylenes by meta-stable iron sulfides and Zn(0)
(35, 37). However, the degradation rate by pyrite decreased
as the reaction proceeded and the reductive capacity of pyrite
was consumed. A similar trend on the degradation rate of
TCE by pyrite was observed in Weerasooriya and Dhar-
masena’s experiment (32). The rate constant for the reductive
dechlorination of VC at the reactive surfaces of pyrite has the
greatest value followed by those for TCE, PCE, and cis-DCE,
although the constants are similar to each other differing by
a factor of only 1.7. Over the 70-day sampling period, there
were no significant differences observed for the kinetics of
TCE removal between the samples with and without biocide
(NaN3), which shows that the reductive dechlorination
process in these experiments was abiotic. Specific reductive
capacities of pyrite for target organics were calculated by
The material balance equation in the batch reactor can
be combined with the rate equation to give the following
relationship expressed in terms of aqueous concentrations.
dC
CE ) -
dt
(k/ pCE){C0RC - pCE(CC0 E - CCE)}CCE
1/ K + CCE
(5)
The kinetic parameters (k, C0RC, and K) were obtained by
an optimization procedure using MATLAB (MathWorks Inc.).
This procedure solves eq 5 numerically by a fourth-order
Runge-Kutta method, calculates the sum of squares for the
parameters, and then minimizes it by adjusting values of the
parameters with the Marquardt-Levenberg algorithm.
To compare the dechlorination kinetics reported here to
those reported by others, corrected pseudo-first-order initial
rate constants (k1, day-1) were calculated using eq 6, which
was derived from equations 1 and 5.
(k/ pCE)C0RC
1/ K + C0CE
k1 )
(6)
k1 was then normalized by the surface area of the soil minerals
(27.8 m2/ g (pyrite) and 57.2 m2/ g (magnetite)) (40) to give a
surface area-normalized pseudo-first-order initial rate con-
stant (k1,sa, Lm-2 day-1).
Reductive Dechlorination of Chlorinated Ethylenes by
Pyrite. Figure 1 shows the decay of target organics in controls
and pyrite suspensions. TCE concentration in controls
abruptly decreased at the start of the experiment due to
sorption on the solid phases of the reactor system (Teflon
septum liner + reactor wall). It reached approximately 97%
9
VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 5 1 4 9

TABLE 1. Kinetic Parameters, Transformation Products and Their Recoveries, and Target Organica Remaining in Pyrite and
Magnetite Suspensionsb at the Last Sampling Time
soil
minerals
target
organics
specific reductive
product recovery and
target organic remaining (%)
capacity (µM/g)c
K (mM-1
)
k (day-1
)
R2 d
pyrite
PCE
1.01 ((0.01)g
0.642 ((0.010)
1.01 ((0.02)
0.974
TCE: 6.7
C2H2: 32.0
C2H4: 6.0
PCE: 55
totale: 99.7
cis-DCE: 3.3
C2H2: 43.0
TCE
1.48 ((0.03)
1.47 ((0.04)h
0.345 ((0.011)
0.346 ((0.010)h
1.59 ((0.02)
1.60 ((0.02)h
0.988
0.989h
C2H4: 2.2
TCE: 50
totale: 98.5
C2H2: 13.2
C2H4: 5.2
cis-DCE: 70.2
totale: 88.6
C2H4: 34
cis-DCE
VC
1.49 ((0.04)
2.26 ((0.06)
0.300 ((0.005)
0.187 ((0.006)
0.984 ((0.021)
1.71 ((0.06)
0.983
0.970
C2H6: 5.0
VC: 56
totale: 95
m agnetite
PCE
0.33 ((0.02)
0.37 ((0.02)
0.700 ((0.022)
0.503 ((0.014)
0.202((0.017)
0.254 ((0.010)
0.910
0.961
chloride: 5.5
PCE: 90.1
totalf: 95.6
chloride: 10.7
TCE: 86.3
TCE
totalf: 97
cis-DCE
VC
0.54 ((0.06)
2.06 ((0.16)i
0.501 ((0.056)
0.344 ((0.018)i
0.185 ((0.023)
0.28 ((0.023)i
0.817
chloride: 1.95/28.7i
cis-DCE: 96.1/61.3i
totalf: 98.1/90i
chloride: 4.2/43.9i
VC: 93.7/48.3i
totalf: 97.9/92.2i
0.872i
0.73 ((0.07)
2.86 ((0.09)i
0.3 ((0.028)
0.2 ((0.010)i
0.193 ((0.021)
0.35 ((0.019)i
0.823
0.943i
a
Initial concentrations of target organics were 0.19 (PCE), 0.25 (TCE), 0.41 (cis-DCE), and 0.40 m M (VC). pH of soil m ineral suspension was
b
c
constant at 8 (pyrite) and 7 (m agnetite). Mass ratios of solid to water were 0.084 (pyrite) and 0.063 (m agnetite). Specific reductive capacity was
obtained by dividing C0RC by the solid concentration. R2 values of nonlinear regression for kinetic param eters. Total is total carbon m ass balance
d
e
for pyrite. f Total is total chlorine m ass balance for m agnetite. Uncertainties represent 95% confidence lim its. 20 m g NaN3 was added. i Fe(II)
g
h
was added. [Fe(II)] ) 42.6 m M.
TABLE 2. Surface Area-Normalized Pseudo-First-Order Initial Rate Constants (k1,sa)a for the Reductive Dechlorination of
Chlorinated Ethylenes by a Variety of Reductants
reductants (surface
area concn: m2/L)
PCE
TCE
cis-DCE
VC
reference
pyriteb
(2340)
1.97 × 10-5
8.38 × 10-7
2.53 × 10-5
1.32 × 10-5
2.27 × 10-5
this study
m agnetiteb
(3600)
7.21 × 10-7
4.56 × 10-4
0.075 × 10-2
7.15 × 10-2
5.52 × 10-2 g
5.60 × 10-7
5.64 × 10-7
this study
(35)
5.74 × 10-6 h
5.78 × 10-6 h
m agnetitec
(81)
Zn(0)d
8.28
8.45 × 10-5
2.40 × 10-3
(45)
(5.47)
m ackinawitee
(0.5)
2.74 × 10-2
(37)
troilitef
(0.5)
(35)
a
b
Unit of surface area-norm alized pseudo-first-order initial rate constant (Lm -2 day-1). Initial target organic concentration: 0.19 (PCE), 0.25
c
(TCE), 0.41 (cis-DCE), and 0.40 m M (VC), pH of soil m ineral suspension: 7 (m agnetite) and 8 (pyrite). Experim ental condition is not available.
d
Initial target organic concentration: 0.015 to 0.35 m M (PCE, TCE, and cis-DCE), 0.003 to 0.03 m M (VC), pH of Zn(0) suspension (50 m M Tris buffer):
7.2. Initial target organic concentration: 0.013 (PCE) and 0.015 µM (TCE), pH of m ackinawite suspension (50 m M Tris buffer): 8.3. f Initial concentration
e
g
is not available, pH: 7.5 (influent) and 7.3 (effluent). The surface area-norm alized initial rate constant by troilite was calculated assum ing surface
area concentration of troilite is 0.5 m 2/L. Fe(II) was added. [Fe(II)] ) 42.6 m M.
h
dividing the initial concentration of reductive capacity by
the concentration of pyrite. Except for the specific reductive
capacity for VC, they were similar to those measured
previously, differing at most by a factor of 1.5 (40). The 95%
confidence intervals for the kinetic parameters are less than
4%, and the R2 values for the nonlinear regressions are greater
than 0.97. These results indicate that the model reasonably
describes the dechlorination kinetics in pyrite suspensions.
The surface area-normalized pseudo-first-order initial rate
constants are shown in Table 2. Rate constants for TCE were
the highest, followed by those for VC, PCE, and cis-DCE. This
result is interesting because the calculated reduction po-
tentials for one (or two) electron reduction of PCE and cis-
DCE are greater than those of TCE and VC, respectively (45,
47), which indicates that PCE and cis-DCE should be more
susceptible to reductive dechlorination. Similar results for
9
5 1 5 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002

relative rates have been reported for the reductive de-
chlorination of chlorinated ethylenes by mackinawite (37)
and zerovalent metals (45, 48). cis-DCE has the smallest value,
which agrees with previous research (45). The rate constants
for pyrite are smaller than those reported for other reductants.
They are approximately 2-5 orders of magnitude smaller
than those reported for the reductive transformation of target
organics by Zn(0) (45) and was 1390-2900 times smaller
than those reported for mackinawite and troilite (35, 37).
These results indicate that pyrite is less reactive than meta-
stable iron sulfides and Zn(0) for the reductive dechlorination
of chlorinated ethylenes in natural and engineered systems.
A study in a model iron-reducing culture showed that abiotic
reductive dechlorination of chlorinated organic at magnetite
surfaces was 60-260 times faster than biotic reductive
dechlorination using the produced total protein and mineral
surface area during the growth in the system as a basis for
comparison (49). This suggests that pyrite could be more
important than microorganisms under some conditions (e.g.,
sulfate reducing condition) in affecting the fate of chlorinated
ethylenes. The experimental efforts to further investigate the
contribution of abiotic dechlorination process by iron sulfide
minerals in natural reducing environments are needed to
support the hypothesis.
The principal transformation product of VC in pyrite
suspension was ethylene, which accounted for 77.3% of VC
that was removed. Ethylene was further reduced to ethane
in the suspension, accounting for 11.4% of the target organic
removed. Acetylene was not observed throughout the
experiment. A similar result was found in abiotic reductive
transformation of VC by Zn(0), but no reduction to ethane
was reported for Zn(0) (45). Based on these results, a
sequential hydrogenolysis pathway can be hypothesized to
explain the reductive transformation of VC in pyrite suspen-
sion (VC f ethylene f ethane).
The reductive dechlorination of cis-DCE in pyrite sus-
pension produced acetylene and ethylene, which accounted
for 44.3% and 17.5% of removed cis-DCE, respectively. Neither
VC nor ethane was observed above the method detection
limit (VC: 16 µM and ethane: 33.2 µM), while a trace amount
of VC was detected as a transformation product of cis-DCE
and trans-DCE by zerovalent Fe(0) and Zn(0) (45, 47). Based
on the distribution of transformation products, the reductive
elimination pathway is suggested to be a main pathway for
the reductive dechlorination of cis-DCE by pyrite. Proposed
pathways for reductive dechlorination of chlorinated eth-
ylenes are shown in Figure 2.
The main transformation product for TCE dechlorination
was acetylene, accounting for 86% of TCE that was removed.
Minor products were cis-DCE (6.6% of removed TCE) and
ethylene (4.4% of removed TCE). No 1,1-DCE was observed
in these experiments, in contrast to those of Weerasooriya
and Dharmasena’s (32). Similar product distributions have
been found during reductive dechlorination of TCE by
mackinawite and troilite, where approximately 70 to 80% of
removed TCE was transformed to acetylene, 7% was trans-
formed to cis-DCE, and 15 to 20% was transformed to
ethylene, ethane, and other C2 to C6 hydrocarbon (35, 37).
Because acetylene was the main transformation product, a
reductive elimination pathway would appear to be the
primary one. Acetylene could also be produced by first
converting TCE to cis-DCE. If this were an important pathway,
the yield of acetylene relative to cis-DCE would increase over
time. However, this yield was observed to be almost constant,
so this pathway is probably not important. Removal of TCE
without observing stoichiometric production of a chlorinated
intermediate suggests that the acetylene formed during the
TCE dechlorination is produced via an unstable intermediate
(chloroacetylene) that quickly decays. One possible reaction
pathway, TCE f chloroacetylene f VC f ethylene, has been
FIGURE 2. Possible pathways for the reductive dechlorination of
chlorinated ethylenes by pyrite. Chemical compounds in bold
characters were detected in the experiments (see Figure S-2,
Supporting Information).
proposed for dechlorination of TCE by Zn(0) (45). However,
this reaction does not seem to apply to the transformation
of TCE in pyrite suspension because of the absence of VC.
Therefore, TCE f chloroacetylene f acetylene f ethylene
would be a reasonable main transformation pathway (see
Figure S-2, Supporting Information).
The main transformation product of PCE dechlorination
by pyrite was acetylene, which accounted for 71.1% of
removed PCE. Minor products were ethylene (13.3% of
removed PCE) and TCE (14.9% of removed PCE). Neither
cis-DCE nor VC was observed above the method detection
limit (cis-DCE: 2.06 µM and VC: 16 µM). PCE could be
transformed to TCE, followed by the parallel transformation
to acetylene and cis-DCE. Because cis-DCE was not detected
during the transformation of PCE, the reaction pathways,
PCE f TCE f cis-DCE f acetylene and PCE f TCE f cis-
DCE f VC, are not supported by observations. If TCE were
a main intermediate for the transformation of PCE to
acetylene, the yield of acetylene relative to TCE would increase
over time. However the yield was almost constant, suggesting
that other parallel transformation pathways with highly
reactive intermediates (dichloroacetylene and chloroacety-
lene) exist for the transformation of PCE to acetylene. Because
no DCEs were observed during the reaction, the reductive
transformation of PCE to acetylene seems to follow the
pathway of PCE f dichloroacetylene f chloroacetylene f
acetylene rather than the pathway of PCE fdichloroacetylene
f cis-DCE or trans-DCE f acetylene.
Reductive Dechlorination of Chlorinated Ethylenes by
Magnetite. Figure 3 shows the decay of target organics in
controls and magnetite suspension with and without addition
of 42.6 mM Fe(II). Concentrations are shown as total chlorine
concentration (organic chlorine and chloride) relative to the
9
VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 5 1 5 1

et al. (35, 50). The greater rate constant reported by Sivavec
et al. could be caused by the lower initial concentration of
TCE used in their experiments if saturation kinetics apply
rather than first-order kinetics.
Ferrous iron was added to investigate the effect of
additional Fe(II) on the dechlorination kinetics of cis-DCE
and VC. Such additional Fe(II) could be produced by the
microbial reduction of Fe(III) to Fe(II), which would provide
a continuous source of reducing power in natural environ-
ments. No significant difference was observed on the decay
of cis-DCE and VC between the controls with and without
Fe(II) addition (see Figure S-3, Supporting Information)
indicating that the reductive dechlorination of target organics
by dissolved Fe(II) is not important. No obvious precipitates
were observed in the controls with Fe(II), although a very
thin layer of precipitate was occasionally observed in the
controls (after 60 days) after centrifugation. No analytical
effort was made to identify the precipitates, and no significant
difference was observed on the dechlorination kinetics
between the controls with and without the precipitate. This
indicates that the precipitate (presumably, siderite (FeCO3))
did not play a significant role in the dechlorination of target
organics. The addition of Fe(II) to magnetite suspensions
increased the dechlorination rates of cis-DCE and VC by a
factor of 10 and increased the specific reductive capacity of
magnetite for cis-DCE and VC by a factor of approximately
4. The increased reactivity of magnetite caused by Fe(II)
addition may be due mainly to the formation of reactive
surface-bound Fe(II) species. The reactive surface-bound Fe-
(II) species responsible for the increased decay rate would
be surface precipitates or surface clusters rather than isolated
surface complexes of Fe(II) such as tFeOFe+ or tFeOFeOH0
because the high Fe(II) concentration and sufficient Fe(II)
contact time employed in this experiment could allow for
the rearrangement of the surface complexes to more reactive
surface precipitate or cluster forms (51). The enhanced
reactivity of magnetite due to Fe(II) addition indirectly shows
the potential link between biotic and abiotic processes in
the reductive dechlorination of cis-DCE and VC in contami-
nated plumes. Microorganisms (e.g., Geobacter metallire-
ducens) can reduce hydrous ferric oxide to magnetite (39,
49) and/ or to Fe(II) surface species bound to the magnetite
under Fe(III) reducing conditions. These biogenic magnetite
and Fe(II) surface species could be regenerated by Fe(II)
produced by microorganisms after their oxidation by chlo-
rinated organics. The effect of Fe(II) addition to the magnetite
suspension, therefore, may be due to the reductive regen-
eration of magnetite by Fe(II) or by the activity of Fe(II) surface
species bound to magnetite or its oxidation products (38,
52).
The results obtained in these laboratory studies can be
extrapolated to estimate behavior in saturated soils and
aquifers using eq 6. These calculations are presented to
demonstrate the potential for abiotic processes to impact
the fate of chlorinated organics, not to predict actual behavior.
The initial reductive capacity concentration (C0RC) in the field
can be calculated by assuming that it is proportional to the
concentration of iron. The concentration of iron (mass Fe/
volume water) in the field is calculated as 91 g/ L by assuming
an iron content of 2.6%, bulk density of 1.4 kg/ L, and porosity
of 0.40. The concentration in the laboratory experiments was
calculated as 18.1 and 45.5 g/ L for pyrite and magnetite
suspensions, respectively, using their solid/ liquid mass ratios
(0.084 and 0.063, respectively) and their measured iron
contents (0.215 g/ g and 0.722 g/ g, respectively). A partitioning
factor (pCE) for PCE in the field of 4.6 can be calculated by
further assuming a soil organic fraction of 0.005 and an
organic carbon partition coefficient of 206 L/ kg. Kinetic
coefficients for assumed field conditions were calculated by
applying results from Table 1 and are shown in Table 3. The
FIGURE 3. Reductive dechlorination of chlorinated ethylenes in
magnetite suspension (0.063 g/g) with and without 42.6 mM Fe(II)
addition. Curves represent kinetic model fits. Control is the control
sample for PCE.
initial total chlorine concentration. The concentration of PCE
in controls rapidly dropped due to the partitioning to the
solid phases of the reactor system and reached equilibrium
concentration (approximately 96% of initial PCE concentra-
tion) in 3 days. The concentrations of controls for TCE, cis-
DCE, and VC (not shown) also decreased rapidly and reached
equilibrium (96.5, 97.7, and 98%, respectively) in 3 days. The
rates of target organics partitioning to solid phases in controls
and magnetite suspensions were also much faster than the
dechlorination rates of target organics by magnetite. The
amount of sorption of each target compound in magnetite
suspension was greater than that in pyrite suspension. The
surface area of magnetite is 2.1 times greater than that of
pyrite, which may cause higher sorption capacity of magnetite
for target organics.
The trends in decay of target organics observed in
magnetite suspension with and without Fe(II) were similar
to those observed in pyrite suspensions. Approximately, 3.9
to 13.7% of the initial target organics were removed in
magnetite suspension and 1.95 to 10.7% were dechlorinated
in 100 days. The relative total chlorine concentration of target
organics (chlorine balance) in magnetite suspension was 95.6
to 98.1% at the last sampling time. The rate constant for the
reductive dechlorination of TCE at the reactive magnetite
surfaces is the greatest, followed by those for PCE, VC, and
cis-DCE (Table 1). The specific reductive capacity of magnetite
for target organics shown in Table 1 was 2.8 to 4 times smaller
than that of pyrite, which is very similar to the measured
results previously reported by Lee (40). The sorption coef-
ficients of target organics in magnetite suspension were
slightly greater than those in pyrite suspension by a factor
of 1.1 to 1.7. The k1,sa for the reductive dechlorination of PCE
by magnetite (Table 2) has the greatest value followed by
those for TCE, VC, and cis-DCE, in contrast to the results for
pyrite suspensions in which TCE dechlorination was most
rapid. The surface area-normalized rate constants for the
dechlorination of different target organics by pyrite and
magnetite are similar for each soil mineral, differing by a
factor of less than 2. This result provides evidence that an
electron transfer may not be the rate-limiting step in these
reactions. However, no efforts were made to identify another
type of controlling reaction mechanisms in this research.
The reductive dechlorination of target organics by magnetite
is much slower than dechlorination by other reductants. For
example, values of k1,sa for magnetite were approximately
23.5 to 40.3 times smaller than those for pyrite and 2-7 orders
of magnitude smaller than those for Zn(0) (45). The k1,sa for
reductive dechlorination of TCE by magnetite observed in
this study was 630 times smaller than that reported by Sivavec
9
5 1 5 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002

(14) Holliger, C.; Schumacher, W. Antonie Van Leeuwenhoek 1994,
66, 239-246.
TABLE 3. Extrapolated Kinetic Parameters for PCE Degradation
under Field Conditions
(15) Gossett, J. M. Microbiological Aspects Relevant to Natural
Attenuation of Chlorinated Ethenes. Ward, C. H., Ed.; In
Symposium on Natural Attenuation of Chlorinated Organics in
Ground Water; EPA: Dallas, TX, 1996; pp 10-13.
(16) McCarty, P. L. Biotic and Abiotic Transformations of Chlorinated
Solvents in Ground Water. Ward, C. H., Ed.; In Symposium on
Natural Attenuation of Chlorinated Organics in Ground Water;
EPA: Dallas, TX, 1996; pp 5-9.
0
reductants
CRC,F (mM/L)a
k1 (day-1
)
t1/2 (days)b
pyrite
m agnetite
0.428
0.042
0.054
13
608
1.1 × 10-3
a
b
Initial reductive capacity concentration in the field. Degradation
half-life.
(17) Butler, E. C. Ph.D. Dissertation, University of Michigan: Ann
Arbor, MI, 1998.
half-life for PCE degradation by pyrite under field conditions
is smaller than often observed in the field, which indicates
that abiotic degradation by pyrite could be an important
factor. The half-life for PCE degradation by magnetite is much
larger but could be reduced if Fe(II) were formed biologically.
These results indicate that abiotic reductive dechlorination
may be an important factor in determining fate of chlorinated
organics during natural attenuation.
(18) Roberts, A. L.; Sanborn, P. N.; Gschwend, P. M. Environ. Sci.
Technol. 1992, 26, 2263-2274.
(19) Schwarzenbach, R. P.; Giger, W.; Schaffner, C.; Wanner, O.
Environ. Sci. Technol. 1985, 19, 322-327.
(20) Hagg, W. R.; Mill, T. Environ. Toxicol. Chem. 1988, 7, 917-924.
(21) Barbash, J. E. Ph.D. Dissertation, Stanford University: Stanford,
CA, 1993.
(22) Kriegman-King, M. R.; Reinhard, M. Environ. Sci. Technol. 1992,
26, 2198-2206.
(23) Curtis, G. P.; Reinhard, M. Environ. Sci. Technol. 1994, 28, 2393-
Acknowledgments
2401.
(24) Dunnivant, F. M.; Schwarzenbach, R. P.; Macalady, D. L. Environ.
Sci. Technol. 1992, 26, 2133-2141.
(25) Schwarzenbach, R. P.; Stierli, R.; Lanz, K.; Zeyer, J. Environ. Sci.
Technol. 1990, 24, 1566-1574.
(26) Burris, D. R.; Delcomyn, C. A.; Smith, M. H.; Roberts, A. L.
Environ. Sci. Technol. 1996, 30, 3047-3052.
(27) Assaf-anid, N.; Hayes, K. F.; Vogel, T. M. Environ. Sci. Technol.
1994, 28, 246-252.
This research has been funded entirely with funds from the
State of Texas as part of the program of the Texas Hazardous
Waste Research Center. The contents do not necessarily
reflect the views and policies of the sponsor nor does the
mention of trade names or commercial products constitute
endorsement or recommendation for use.
(28) Chiu, P.; Reinhard, M. Environ. Sci. Technol. 1995, 29, 595-603.
(29) Chiu, P.; Reinhard, M. Environ. Sci. Technol. 1996, 30, 1882-
1889.
(30) Schreier, C. G. M.S. Thesis, Stanford University: Stanford, CA,
1990.
(31) Kriegman-King, M. R.; Reinhard, M. Chapter 16: Reduction of
Hexachloroethane and Carbon Tetrachloride at Surface of
Biotite, Vermiculite, Pyrite, and Marcasite. In Organic Substances
and Sediments in Water; Baker, R. A., Ed.; Lewis Publishers Inc.:
Chelsea, MI, 1991.
Supporting Information Available
Anaerobic environments, preparation of soil minerals, char-
acterization of soil minerals, derivation of kinetic equations,
and solid phase partition coefficients for chlorinated eth-
ylenes, X-ray diffraction patterns of pyrite and magnetite
used in this research (Figure S-1), reductive transformation
of 0.25 mM TCE in 0.084 g/ g pyrite suspension (Figure S-2),
reductive dechlorination of 0.4 mM VC by magnetite (0.063
g/ g) with and without 42.6 mM Fe(II) addition (Figure S-3),
characteristics of soil minerals (Table S-1), and solid phase
partitioning coefficients (ks) and partitioning factors (pCE) of
chlorinated ethylenes in pyrite and magnetite suspensions
(Table S-2). This material is available free of charge via the
Internet at http:/ / pubs.acs.org.
(32) Weerasooriya, R.; Dharmasena, B. Chemosphere 2001, 42, 389-
396.
(33) Klausen, J.; Tro¨ ber, S. P.; Haderlein, S. B.; Schwarzenbach, R.
P. Environ. Sci. Technol. 1995, 29, 2396-2404.
(34) Kriegman-King, M. R. Environ. Sci. Technol. 1994, 28, 692-700.
(35) Sivavec, T. M.; Horney, D. P. Reduction of Chlorinated Solvents
by Fe(II) Minerals. 213th ACS National Meeting; ACS, 1997; pp
115-117.
Literature Cited
(36) Butler, E. C.; Hayes, K. F. Environ. Sci. Technol. 1998, 32, 1276-
1284.
(1) McCarty, P. L.; Semprini, L. Handbook of Bioremediation; Lewis
Publishers: Boca Raton, FL, 1994; pp 87-116.
(2) Bouwer, E. J. Handbook of Bioremediation; Lewis Publishers:
Boca Raton, FL, 1994; pp 149-175.
(3) Weaver, J. W.; Wilson, J. T.; Kampbell, D. H. Case Study of Natural
Attenuation of Trichloroethene at St. Joseph, Michigan. Ward,
C. H., Ed.; In Symposium on Natural Attenuation of Chlorinated
Organics in Ground Water; EPA: Dallas, TX, 1996; pp 69-73.
(4) EPA U.S. Common Chemicals Found at Superfund Sites; EPA
540/ R-94/ 044; Office of Emergency and Remedial Response:
Washington, DC, 1994.
(5) Leisinger, T. In Biotechnology; Rehm, H. J., Reed, G., Eds.; VCH
Verlagsgesellschaft: Weinheim, 1986; Vol. 8, pp 475-513.
(6) Hutzinger, O.; Veerkamp, W. In Microbial Degradation of
Xenobiotic and Recalcitrant Compounds; Leisinger, T., Ed.;
Academic Press: London, 1981.
(7) Hileman, B. Chem. Eng. News 1993, 19, 11-20.
(8) Ukrainczyk, L.; Chibwe, M.; Pinnavaia, T. J.; Boyd, S. A. Environ.
Sci. Technol. 1995, 29, 439-445.
(9) Bagley, D. M.; Gossett, J. M. Appl. Environ. Microbiol. 1990, 56,
2511-2516.
(10) Freedman, D. L.; Gossett, J. M. Appl. Environ. Microbiol. 1989,
55, 2144-2151.
(11) Fathepure, B. Z.; Nengu, J. P.; Boyd, S. A. Appl. Environ. Microbiol.
1987, 53, 2671-2674.
(12) Fathepure, B. Z.; Boyd, S. A. Appl. Environ. Microbiol. 1988, 54,
2976-2980.
(37) Butler, E. C.; Hayes, K. F. Environ. Sci. Technol. 1999, 33, 2021-
2027.
(38) Lee, W.; Batchelor, B. Abiotic Reductive Dechlorination of
Chlorinated Ethylenes by Iron-bearing Soil Minerals and
Potential Interactions with Biotic Processes. In Chemical-
Biological Interactions in Contaminant Fate; Tratnyek, P. G.,
Adriaens, P., Roden, E. E., Eds.; 220th ACS National Meeting;
American Chemical Society: Washington, DC, 2000; pp 338-
340.
(39) McCormick, M. L.; Kim, H. S.; Adriaens, P. Transformation of
Tetrachloromethane in
a Defined Iron Reducing Culture:
Relative Contributions of Cell and Mineral Mediated Reactions.
In Specialty Chemicals in the Environment; Stone, A. T., Ed.;
219th ACS National Meeting; American Chemical Society: San
Francisco, CA, 2000; pp 138-141.
(40) Lee, W. Ph.D. Dissertation, Texas A&M University: College
Station, TX, 2001.
(41) Taylor, R. M.; Maher, B. A.; Self, P. G. Clay Miner. 1987, 22,
411-422.
(42) Kriegman-King, M. R. Ph.D. Dissertation, Stanford University:
Stanford, CA, 1993.
(43) Hwang, I.; Batchelor, B. Environ. Sci. Technol. 2000, 34, 5017-
5022.
(44) Gossett, J. M. Environ. Sci. Technol. 1987, 21, 202-208.
(45) Arnold, W. A.; Roberts, A. L. Environ. Sci. Technol. 1998, 32,
3017-3025.
(13) Holliger, C.; Schraa, G.; Stams, A. J. M.; Zehnder, A. J. B. Appl.
Environ. Microbiol. 1993, 59, 2991-2997.
(46) Mackay, D.; Shiu, W. Y. J. Phys. Chem. Ref. Data 1981, 10, 1175-
1199.
9
VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 5 1 5 3

(47) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell,
T. J. Environ. Sci. Technol. 1996, 30, 2654-2659.
(48) Gillham, R. W.; O’Hannesin, S. F. Ground Water 1994, 32, 958-
967.
(51) Pecher, K.; Haderlein, S. B.; Schwarzenbach, R. P. Environ. Sci.
Technol. 2002, 36, 1734-1741.
(52) Heijman, C. G.; Grieder, E.; Holliger, C.; Schwarzenbach, R. P.
Environ. Sci. Technol. 1995, 29, 775-783.
(49) McCormick, M. L.; Bouwer, E. J.; Adriaens, P. Environ. Sci.
Technol. 2002, 36, 403-410.
(50) Sivavec, T. M.; Park, C. Inventors; General Electric Company,
Assignee; Composition and Method for Ground water Reme-
diation. USA patent 5750036. May, 1998.
Received for review May 30, 2002. Revised manuscript re-
ceived September 13, 2002. Accepted September 25, 2002.
ES025836B
9
5 1 5 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002