Factors influencing rates and products in the transformation of trichloroethylene by iron sulfide and iron metal

Article abstract of DOI:10.1021/es010620f

Batch experiments were performed to assess (i) the influence of pH, solution amendments, and mineral aging on the rates and products of trichloroethylene (TCE) transformation by iron sulfide (FeS) and (ii) the influence of pretreatment of iron metal with NaHs on TCE transformation rates. The relative rates of FeS-mediated transformation of TCE to different products were quantified by branching ratios. Both pseudo-first-order rate constants and branching ratios for TCE transformation by FeS were significantly influenced by pH, possibly due to a decrease in the reduction potential of reactive surface species with increasing pH. Neither Mn²⁺, expected to adsorb to FeS surface S atoms, nor 2,2′-bipyridine, expected to adsorb to surface Fe atoms, significantly influenced rate constants or branching ratios. FeS that had been aged at 76 °C for 3 days was completely unreactive with respect to TCE over 6.5 months, yet this aged FeS transformed hexachloroethane to tetrachloroethylene with a rate constant only slightly lower than that for nonaged FeS. This finding suggests that the oxidation state of iron sulfide minerals in the environment will strongly influence the potential for intrinsic remediation of pollutants such as TCE. Treatment of iron metal with bisulfide significantly increased the pseudo-first-order rate constant for TCE transformation at pH 8.3. This effect was attributed to formation of a reactive FeS coating or precipitate on the iron surface.

Full text of DOI:10.1021/es010620f

Environ. Sci. Technol. 2001, 35, 3884-3891
Factors Influencing Rates and
Products in the Transformation of
Trichloroethylene by Iron Sulfide
and Iron Metal
E L I Z A B E T H C . B U T L E R * A N D
K I M F . H A YE S
Department of Civil and Environmental Engineering,
University of Michigan, Ann Arbor, Michigan 48109-2125
FIGURE 1. Pathways for TCE transformation by FeS (16, 17, 38-45).
rates. TCE was chosen for study since it is a common
groundwater pollutant that has been detected at more than
half of all National Priorities List (NPL or Superfund) sites in
the United States (2), and because its transformation leads
to reaction products that vary significantly in their human
toxicity.
Batch experiments were performed to assess (i) the
influence of pH, solution amendments, and mineral aging
on the rates and products of trichloroethylene (TCE)
transformation by iron sulfide (FeS) and (ii) the influence
of pretreatment of iron metal with NaHS on TCE transformation
rates. The relative rates of FeS-mediated transformation
of TCE to different products were quantified by branching
ratios. Both pseudo-first-order rate constants and
branching ratios for TCE transformation by FeS were
significantly influenced by pH, possibly due to a decrease
in the reduction potential of reactive surface species
with increasing pH. Neither Mn2+, expected to adsorb to
FeS surface S atoms, nor 2,2′-bipyridine, expected to adsorb
to surface Fe atoms, significantly influenced rate constants
or branching ratios. FeS that had been aged at 76 °C
for 3 days was completely unreactive with respect to TCE
over 6.5 months, yet this aged FeS transformed hexachloro-
ethane to tetrachloroethylene with a rate constant only
slightly lower than that for nonaged FeS. This finding suggests
that the oxidation state of iron sulfide minerals in the
environment will strongly influence the potential for intrinsic
remediation of pollutants such as TCE. Treatment of iron
metal with bisulfide significantly increased the pseudo-first-
order rate constant for TCE transformation at pH 8.3.
This effect was attributed to formation of a reactive FeS
coating or precipitate on the iron surface.
Corrosion of zerovalent iron metal and steel by aqueous
sulfide species results in the formation of iron sulfide minerals
(3-14). At neutral to alkaline pH values, the initial iron sulfide
phase formed is highly crystalline mackinawite (FeS1-x) (4,
7, 15). Mackinawite has also been observed to form from the
in situ reaction of iron trash with sulfide species formed by
sulfate-reducing bacteria in the Mystic River, Boston, MA (4,
5) and from exposure of a steel coupon to H2S producing
bacteria (3). Iron sulfide minerals such as mackinawite are
reactive in the transformation of halogenated organic pol-
lutants (16-26), and treatment of iron granules with sulfide
has been shown to increase the rate of transformation of
TCE (27, 28), tetrachloroethylene (PCE) (28), and carbon
tetrachloride (CT) (16, 29). FeS has been reported to be
significantly more reactive per unit surface area than iron
metal in TCE transformation (21). Together, these results
suggest that iron metal placed in subsurface reactive barriers
could naturally (i.e., through the growth of sulfate-reducing
bacteria) or through engineered measures (e.g., injection of
sulfide) form an FeS coating that could enhance its reactivity
with pollutants such as TCE.
Mackinawite is also the initial iron sulfide mineral formed
from the reaction of naturally occurring iron minerals such
as goethite (R-FeOOH) with aqueous sulfide species (5, 15),
although the mackinawite formed through this process is
typically poorly crystalline (e.g., ref 15). Mackinawite may
age to other iron sulfide minerals such as greigite (Fe3S4) (6,
15, 30, 31) and pyrite (FeS2) (6, 31). Although these phase
transformations are thermodynamically favorable (14) in the
presence of oxidants such as polysulfides (15), elemental
sulfur (6, 31), or oxygen (6, 15, 32), aging rates are dependent
on solution pH (15), temperature (15, 33), and surface area
of the solid phase (15, 33) and may be on the order of months
(34). Consequently, the metastable phases mackinawite and
greigite may persist in the environment under ambient
conditions (6, 15) and thus may be important in natural or
engineered in situ remediation technologies. Both macki-
nawite (4, 5, 35-37) and greigite (38) have been identified
in natural systems.
Introduction
Use of reactive minerals and metals in the transformation
of chlorinated organic pollutants has shown promise in
remediation strategies such as natural attenuation and
subsurface reactive barriers (e.g., ref 1). The success of such
remediation strategies depends on overall pollutant disap-
pearance rates and the extent that pollutants are transformed
to less harmful products. The goal of this work was to provide
quantitative data on the rates and products of trichloro-
ethylene (TCE) transformation by iron sulfide (FeS) and by
iron metal treated with NaHS. Experiments were performed
to assess (i) the influence of pH, solution amendments, and
mineral aging on the rates and products of trichloroethylene
(TCE) transformation by FeS and (ii) the influence of
pretreatment of iron metal with NaHS on TCE transformation
TCE can undergo reductive transformation by at least
two pathways in the presence of FeS (21, 23), as illustrated
schematically in Figure 1. Additional reaction pathways, not
shown in Figure 1, are important in the transformation of
TCE by iron metal (17, 39, 40). In pathway 1, TCE undergoes
dichloroelimination to acetylene via the transient intermedi-
ate chloroacetylene (39-44). Acetylene may undergo sub-
* Corresponding author phone: (405)325-3606; fax; (405)325-4217;
e-mail: ecbutler@ou.edu. Present address: School of Civil Engineer-
ing and Environmental Science, University of Oklahoma, Norman,
OK, 73019-0631.
9
3 8 8 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001
10.1021/es010620f CCC: $20.00
2001 Am erican Chem ical Society
Published on Web 08/28/2001

sequent hydrogenation to ethylene and/ or ethane under
certain conditions (e.g., refs 21, 39, 40, 45, 46). In pathway
2, TCE undergoes sequential hydrogenolysis, forming cis-
1,2-dichloroethylene (cis-DCE), vinyl chloride (VC), and
ethylene. Sequential hydrogenolysis may result in accumu-
lation of the harmful products cis-DCE and VC, both regulated
under the U.S. Safe Drinking Water Act, so understanding
the factors that favor dichloroelimination versus hydro-
genolysis can lead to remediation technologies that do not
result in production of cis-DCE or VC at concentrations above
regulatory limits. While previously reported studies of TCE
transformation by FeS (17, 21, 23), pyrite (FeS2) (47, 48), and
iron metal (17, 39, 40) found that TCE transformation takes
place primarily by pathway 1, these studies did not focus on
how variation of system properties could affect the distribu-
tion of TCE reaction products, one focus of this research.
ampules to pop or explode. Solution pH, metal loading, and
the amount of headspace in the ampules can all influence
the potential for pressurization of H2 gas. Therefore, readers
should take these factors into account and use all appropriate
safety precautions when handling sealed ampules containing
iron or other metals and water. For the experimental
conditions reported here, there was no buildup of excessive
H2 gas in the sealed ampules.
All samples contained either 10 g/ L of FeS, 10 g/ L of aged
FeS, or 100 g/ L of iron metal. In all samples, the pH was
buffered with tris(hydroxymethyl)aminomethane (Tris) and
Tris-HCl at a total concentration (acid plus conjugate base)
of 0.1 M. Tris buffer was chosen for these experiments in part
because of its negligible tendency to form complexes with
transition metals (51), making it unlikely to adsorb to a
significant extent to the FeS or iron metal surface. Solution
pH was adjusted to the desired value by addition of small
amounts of HCl or NaOH. Ionic strength was adjusted to 0.1
M by addition of NaCl. All rate constants reported here are
pseudo-first-order rate constants for these conditions. Trans-
formation of TCE was monitored over the course of several
half-lives, except for one experiment with aged FeS for which
no reaction was observed over the course of 6.5 months.
There was no disappearance of TCE in the ionic medium
alone (i.e., only Tris buffer plus NaCl) in the time scale of
these experiments.
Samples were spiked with 50 µL of a 0.002 M solution of
TCE that had been prepared in N2-sparged methanol so that
the resulting aqueous solution contained 1% methanol by
volume. For one experiment, the TCE spiking solution was
prepared in N2-sparged 2-propanol in order to assess the
influence of methanol and 2-propanol on reaction rates and
products. Initial aqueous concentrations of TCE after par-
titioning to the ampule headspace, determined by measure-
ment of the aqueous concentration in the ionic medium
alone, ranged from 15 to 18 µM.
At regular intervals during the course of the reaction,
ampules were centrifuged at approximately 1000 rpm and
broken open, and 100 µL of the supernatant was removed
with a microsyringe and extracted with 0.4 mL of 2,2,4-
trimethylpentane containing 6 µM of 1,3,5-trichlorobenzene
as an internal standard. Extracted samples were analyzed
for TCE using GC method A (23), which employed a Hewlett-
Packard (HP) (Palo Alto, CA) 6890 GC with a J&W Scientific
(Folsom, CA) DB-5 column and an electron capture detector.
Concentrations of TCE were quantified by comparison of
GC peak areas to a five-point standard curve. Samples were
analyzed in duplicate, and the results typically agreed within
1% using GC method A.
A 1 mL volume of the aqueous supernatant from each
sample was also analyzed for cis- and trans-1,2-dichloro-
ethylenes (cis-DCE and trans-DCE), 1,1-dichloroethylene (1,1-
DCE), ethane, ethylene, and acetylene using GC method B
(22), which employed a HP 5890 GC with a HP 19395
headspace autosampler. In GC method B, carrier gas flow
was from the headspace autosampler through a J&W Scientific
DB-624 column and then was split via a valve to a second
DB-624 column in parallel with a J&W Scientific GS-Q column
leading to an ECD and a flame ionization detector (FID),
respectively. Because methanol coeluted with VC, VC could
not be identified or quantified with this method. Concentra-
tions of each compound were quantified by comparison of
GC peak areas to a five-point standard curve. Samples were
analyzed in duplicate, and the results typically agreed within
5% using GC method B. Although GC method B could identify
the presence of each of the compounds ethane, ethylene,
and acetylene, it could not fully resolve the peaks for ethane
and acetylene, so these two compounds could not be
quantified when both were present, which was the case for
all experiments involving iron metal.
Experimental Section
All oxygen-sensitive experimental procedures were con-
ducted in polyethylene glovebags using aqueous reactants
that had been deoxygenated by sparging with high purity N2
(22). Except as noted below, all chemicals were commercially
available reagent or ACS grade and were used as received.
Granular iron metal was from Fisher Scientific (degreased
iron filings, 40 mesh) and Peerless Metal Powders and
Abrasives Company (cast iron aggregate size ETI 8/ 50)
(Detroit, MI). Water was distilled and then purified using a
Milli-Q Plus water system (Millipore Corp., Bedford, MA).
FeS was prepared using a method (22) adapted from
Rickard (34) involving preparation of an FeS slurry by slow
addition of dissolved Na2S to a solution of FeCl2. The FeS was
then freeze-dried and characterized as poorly crystalline
mackinawite with a specific surface area of 0.05 m2/ g (22).
“Aged FeS” was prepared by equilibrating an FeS slurry
prepared in this manner for 3 days in an oven at ap-
proximately 76 °C prior to freeze-drying. The equilibration
took place in serum bottles that were crimp sealed under N2.
While this aging treatment was undertaken in order to
produce mackinawite with a high degree of crystallinity (5,
6, 12), X-ray diffraction analysis of the aged FeS after freeze-
drying indicated only a slightly higher degree of mackinawite
crystallinity versus unaged FeS. In addition to the peaks
characteristic of mackinawite, the X-ray diffraction pattern
for the aged FeS contained a peak, not present in the unaged
FeS, characteristic of greigite (Fe3S4), possibly due to forma-
tion of a greigite surface coating. Unlike the unaged FeS, the
aged FeS was also ferromagnetic, also characteristic of greigite
(5, 6, 50).
Kinetic experiments were conducted in individual 5 mL
flame-sealed glass ampules (21) to maintain anaerobic
conditions and to prevent losses of volatile reactants and
products from individual samples during experiments of up
to 6.5 months duration. After preparation, ampules were
placed on a Labindustries (Berkeley, CA) model T 415-110
rocking platform shaker at approximately 22 cycles per
minute in a temperature-controlled chamber at 25 °C in the
dark. In each ampule, the aqueous phase volume was 5 mL,
and the gas-phase volume was approximately 2.82 mL. The
gas-phase volume was estimated by measuring the average
mass of water required to fill the tops (i.e., above the prescored
necks) of six sealed and broken-open ampules, converting
this mass to a volume, and then adding this to the average
volume required to fill the ampule from the 5 mL level to the
prescored neck. The 95% confidence interval of this average
headspace volume was 0.1105 mL or 3.92%.
Readers should note a safety concern. Because there is
no diffusion of gases into or out of sealed glass ampules,
significant quantities of H2 gas (resulting from metal cor-
rosion) can build up inside sealed ampules containing
zerovalent metals, creating pressures that could cause
9
VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 8 8 5

a
TABLE 1. Observed Products, Mass Recoveries, and Rate Constants for Transformation of TCE by 10 g/L FeS
mass
recovery
(as %)b
kobs (h-1
)
k1 (h-1
)
k2 (h-1
)
{k2′ (h-1)}
conditions
pH 7.3
products
{kobs′ (h-1)}
{k1′ (h-1)}
k1′/k2′
acetylene
cis-DCE
TCE rem aining
totalc
76
11
12
98
4.1((1.4) × 10-4
{5.0((1.7) × 10-4
3.3((1.1) × 10-4
{4.0((1.4) × 10-4
8.0((3.2) × 10-5
{9.9((3.9) × 10-5
4.1 ( 1.0
}
}
}
pH 8.3d
pH 9.3
acetylene
cis-DCE
VC
TCE rem aining
total
65
6
<1
9
1.20((0.12) × 10-3
{1.49((0.14) × 10-3
1.11((0.11) × 10-3
{1.37((0.13) × 10-3
9.4((1.2) × 10-5
11.8 ( 1.1
}
}
}
}
{1.17((0.15) × 10-4
}
80
acetylene
cis-DCE
TCE rem aining
total
96
2
6
1.62((0.35) × 10-3
{2.00((0.43) × 10-3
1.58((0.34) × 10-3
{1.96((0.42) × 10-3
3.91((0.95) × 10-5
40.5 ( 4.7
12.1 ( 1.1
11.38 ( 0.90
9.3 ( 1.4
{4.8((1.2) × 10-5
8.7((1.0) × 10-5
}
104
pH 8.3;
1 m M
MnCl2
acetylene
cis-DCE
TCE rem aining
total
99
7
2
1.137((0.093) × 10-3
1.050((0.086) × 10-3
{1.41((0.12) × 10-3
}
{1.30((0.11) × 10-3
}
{1.07((0.12) × 10-4
}
}
}
108
pH 8.3;
0.1 m M
2,2′-bipyridine TCE rem aining
acetylene
cis-DCE
90
7
3
1.016((0.073) × 10-3
{1.256((0.090) × 10-3
9.34((0.67) × 10-4
8.21((0.84) × 10-5
{1.02((0.10) × 10-4
}
{1.155((0.083) × 10-3
8.5((1.9) × 10-4
}
total
101
pH 8.3;
1% 2-propanol cis-DCE
(by volum e)e
acetylene
79
9
2
9.4((2.1) × 10-4
9.1((2.3) × 10-5
{1.17((0.26) × 10-3
}
{1.05((0.23) × 10-3
}
{1.13((0.29) × 10-4
TCE rem aining
total
90
a
b
c
Uncertainties are 95% confidence intervals. Refer to eq 2 in ref 23 for calculation of m ass recoveries. Sum of products m ay not equal total
d
e
due to round-off error. The values in this row are from ref 23. All other experim ents contained 1% m ethanol (by volum e).
Results and Discussion
TABLE 2. Rate Constants for Transformation of TCE by 100 g/L
Iron Metala
Calculation of Rate Constants. Values of pseudo-first-order
rate constants (kobs values) for TCE disappearance were
determined by nonlinear least squares regression of experi-
mentally measured values of TCE aqueous concentrations
([TCE]aq) versus time, using a pseudo-first-order rate law:
kobs (h-1
)
conditions
{kobs′ (h-1)}
Fisher iron
3.49((0.50) × 10-3
{4.31((0.62) × 10-3
}
}
}
[TCE]aq ) [TCE]aq,0 e-k
(1)
obst
Fisher iron;
1 m M NaHS
6.50((0.59) × 10-3
{8.04((0.73) × 10-3
Experiments were performed in flame-sealed glass am-
pules, which necessarily have some headspace, so there
existed the potential for mass transfer of TCE from the ampule
headspace to the aqueous phase as TCE was removed from
the aqueous phase by reaction with FeS or iron metal. To
eliminate this confounding effect in our kinetic analyses,
values of the pseudo-first-order rate constants that would
be observed in a headspace-free system (kobs′ values) were
calculated using the relationship (23, 43, 52-54)
Peerless iron
1.77((0.51) × 10-3
{2.19((0.64) × 10-3
Peerless iron;
1 m M NaHS
3.78((0.89) × 10-3
{4.7((1.1) × 10-3
}
a
pH 8.3. Uncertainties are 95% confidence intervals. Mass recoveries
are not reported since all products were not quantified.
much more rapidly than TCE transformation reactions, which
is a reasonable assumption given the relatively small rate
constants for the reactions reported here. Values of kobs and
k
obs′ ) kobs fTCE
(2)
k
obs′ are reported in Tables 1 and 2.
The following approach was used to quantify the distri-
Calculation of kobs′ values in this manner allows valid
comparison of rate constants between experimental systems
having different relative headspace volumes. In eq 2 and
below, fi is defined as (1 + Hi × (Vg/ Vaq)), where Vg and Vaq
are the gas and aqueous phase volumes, respectively, and Hi
is the dimensionless Henry’s Law constant for species i, i.e.,
Cg,i/ Caq,i, where Cg and Caq are the gas and aqueous-phase
concentrations, respectively. Dimensionless Henry’s law
constants used in these calculations were averages of all
experimentally determined values at approximately 25 °C
that were reported in refs 55-59, except for Hacetylene, which
was estimated rather than measured (60). These values are
as follows: TCE, 0.419; cis-DCE, 0.167; and acetylene, 0.887.
Application of eq 2 assumes that partitioning of TCE between
the aqueous and gas phases within sealed ampules took place
bution of reaction products in experiments involving FeS
(Table 1). First, eq 3 (23) was used to calculate k1/ k2 or the
branching ratio (61):
k1 [acetylene]aq facetylene
)
(3)
k2
[cis-DCE]aq fcis-DCE
The right-hand side of eq 3 is referred to here as the
acetylene/ cis-DCE product distribution ratio. For each
experiment reported in Table 1, this ratio was calculated at
each sampling time, and the resulting set of product
distribution ratios was averaged. The branching ratio was
then calculated by equating it to the mean acetylene/ cis-
9
3 8 8 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

DCE product distribution ratio. This approach assumed
parallel reaction of TCE to acetylene and cis-DCE and no
other products. Evidence for such a reaction scheme includes
(i) no observation of any trend, either up or down, in the
acetylene/ cis-DCE product distribution ratio for each experi-
ment reported in Table 1 over several half-lives; (ii) the
relatively small 95% confidence intervals for each k1/ k2 value,
and hence each mean acetylene/ cis-DCE product distribution
ratio value, reported in Table 1; and (iii) the nondetection
of any additional reaction products other than minor amounts
of VC after extended time periods (23) (see below). (Formation
of minor amounts of other products that would not be
detected by the analytical methods employed here is also
possible.)
Next, eq 3 was combined with eq 4 (23)
k1 + k2 ) kobs
(4)
FIGURE 2. Aqueous concentrations of TCE, acetylene, and cis-DCE
versus time in the presence of 10 g/L FeS and 0.1 mM 2,2-bipyridine;
pH 8.3. Rate constants for this experiment are reported in the second
to last row of Table 1. Data points represent experimentally measured
values, and lines represent, for [TCE]aq, the solution to eq 1 and, for
[acetylene]aq and [cis-DCE]aq, the solutions to the following equations
to yield eqs 5 and 6
PDR‚kobs
1 + PDR
k1 )
(5)
(6)
(23):
obs,TCEt
kobs
fTCE k1[TCE]aq,0(1 - e-k
)
[acetylene]aq
)
k2 )
facetylene
kobs,TCE
1 + PDR
obs,TCEt
fTCE k2[TCE]aq,0(1 - e-k
)
which were then used to calculate k1 and k2. In eqs 5 and 6,
“PDR” is the mean acetylene/ cis-DCE product distribution
ratio. Finally, values of k1′ and k2′, i.e., those values that would
be measured in a headspace free system, were calculated
using the relationships (23)
[cis-DCE]aq
)
.
fcis-DCE
kobs,TCE
k1′ ) k1 fTCE
(7)
and
k2′ ) k2 fTCE
(8)
Examination of these relationships indicates that k1/ k2 and
k1′/ k2′ are identical. Therefore, only the ratio k1/ k2 is reported
in Table 1.
Reference 23 provides a more detailed development of
this reaction scheme and provides a graph illustrating
reasonable agreement between selected experimental data
and the model developed above. Figure 2 also shows data
from an experiment reported here (Table 1, row 5) along
with model fits, illustrating excellent agreement between data
and model over several half-lives. Figure 2 also illustrates
that acetylene and cis-DCE are not transformed by FeS at
significant rates in the time scale of these experiments.
Evidence reported previously (23) indicates that cis-DCE is
transformed to VC by FeS far more slowly than it is formed
through hydrogenolysis of TCE. In an experiment in which
samples were spiked with a solution of TCE prepared in
2-propanol instead of methanol, VC was not detected until
after approximately 83 days and then at less than 1% of the
original TCE concentration (23).
The 95% confidence intervals for the values reported in
Tables 1 and 2 were calculated by nonlinear regression and/
or propagation of error (62). Details are given in the
Supporting Information.
Transform ation of TCE by FeSsInfluence of pH. Solution
pH had a significant influence on both the overall rate
constants for TCE transformation by FeS (kobs′) as well as the
distribution of reaction products, as shown in Table 1. A
similar dependence of rate constants on pH, although over
a larger pH range, was also observed in the transformation
of TCE by pyrite (48). For the data reported here, eqs 3-8
were used to calculate the relative contributions of k1′ and
k2′ to the overall rate constant, kobs′. Individual values of k1′,
FIGURE 3. k1′, k2′, and kobs′ versus pH for TCE transformation by 10
g/L FeS. Error bars were omitted for clarity. 95% confidence intervals
are reported in Table 1.
k2′, and kobs′ are plotted as functions of pH in Figure 3,
illustrating that the increase in kobs′ observed between pH
7.3 and 9.3 is due primarily to an increase in k1′ or the rate
constant for transformation of TCE to acetylene via pathway
1 (Figure 1). Because k1′ increased significantly between pH
7.3 and 9.3, while k2′ did not change substantially, the
branching ratio increased by a factor of approximately 10
between pH 7.3 and pH 9.3 (Table 1).
Several explanations for the pH-dependence of the overall
rate constant and product distribution were considered.
Figure 4 illustrates a subset of the reactions included in Figure
1 and shows some of the possible elementary reaction steps
involved in the FeS-mediated transformation of TCE to cis-
DCE and acetylene. As illustrated in Figure 4, one possible
pathway for TCE transformation to cis-DCE involves an acid/
base equilibrium between cis-DCE and a dichlorovinyl
carbanion intermediate (Figure 4, species b) (63), possibly
associated with the FeS surface. Assuming that such a
carbanion intermediate approaches equilibrium with its
conjugate acid, cis-DCE, more rapidly than it is either formed
9
VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 8 8 7

FIGURE 4. Possible elementary reaction steps in the transformation of TCE by FeS (16, 17, 38-43, 62). “Nu:” represents a nucleophile.
Species in boxes were identified in these experiments.
from TCE or transformed to chloroacetylene, then lower pH
values would drive the reaction between the carbanion
intermediate and cis-DCE to the right, resulting in higher
concentrations of cis-DCE and lower concentrations of
chloroacetylene and its degradation product acetylene. This
was the rationale used by Seiber (64) to explain the distribu-
tion of hydrogenolysis and dihaloelimination products as a
function of pH in the electrolytic reduction of octachloro-
styrene at a lead cathode.
Another possible pathway for TCE transformation to cis-
DCE involves hydrogen atom abstraction by an adsorbed
radical intermediate such as species (a) in Figure 4. Such a
pathway would be favored under conditions where hydrogen
atom donors are abundant. Hydrogen atoms associated with
FeS surface functional groups such as tFeOH or tFeSH,
where tFe represents a surface iron atom (22), are likely to
be more abundant at low pH, which could explain the pH
dependence of the product distribution. Surface hydride
transfer to an FeS/ TCE organometallic intermediate followed
by chloride ion elimination, resulting in formation of cis-
DCE, would also be favored under conditions where surface
hydrogen atoms are abundant, i.e., at lower pH values. Arnold
and Roberts proposed such a scheme for the transformation
of PCE by iron metal (40).
ering such an acid-base equilibrium, an increase in pH would
result in an increased concentration of deprotonated surface
species. Assuming that the driving force for electron donation
by surface iron atoms is increased by the larger electron
density on more deprotonated ligands (65), these depro-
tonated species would be better reducing agents and thus
could cause an increased rate of reductive dechlorination.
Surface functional groups that are more deprotonated would
also be better nucleophiles, which could increase the rate of
formation of chloroacetylene and its degradation product
acetylene by a nucleophile-induced dichloroelimination
pathway.
In performing cyclic voltammetry with FeS2 (pyrite) and
Fe1-xS (pyrrhotite) electrodes over a wide pH range, Conway
et al. (66) observed an increase in peak currents with
increasing pH. In a controlled-potential technique such as
cyclic voltammetry, the peak current is proportional to the
rate of electron transfer at the electrode-solution interface
(67), so an increase in peak current with increasing pH implies
a faster rate of electron transfer at higher pH values. One
explanation offered by Conway et al. (66) for this pH-
dependence of the peak current was greater deprotonation
of sulfide species in the interfacial region with increasing
pH, an explanation that is generally consistent with the
hypotheses described above. Similarly, Weerasooriya and
Dharmasena (48) explained the influence of pH on the
transformation of TCE by pyrite on pH-dependent equilibria
between pyrite surface functional groups.
Additional insight into factors influencing the branching
ratio in the transformation of TCE by FeS can be gained by
examining the factors influencing the product distribution
in the reductive transformation of CT and other halogenated
methanes in related experimental systems. Halogenated
methanes, like TCE, are susceptible to both hydrogenolysis
(forming haloforms) and dihaloelimination (forming dihalo-
carbenes which degrade rapidly in water to products such
as carbon monoxide (CO) and formate) (68-74). Balko and
Tratnyek (73) studied the reductive transformation of CT by
iron metal in the presence and absence of UV illumination
and attributed the greater concentration of the dihaloelimi-
nation product CO in the UV-illuminated system to the more
negative reduction potential of photogenerated electrons in
the conduction band of the oxide coating on the iron metal.
Consistent with the interpretations discussed above, this
suggests that product distribution can be influenced by the
While these possibilities can explain the effect of pH on
the product distribution in the transformation of TCE by
FeS, they do not explain the increase in overall reaction rate
with increasing pH and therefore cannot entirely explain the
kinetic pattern observed here.
Two other mechanisms might explain the increase in both
k
obs′ and the branching ratio with increasing pH. First, raising
the pH could cause an increase in the rate of formation of
a reactive intermediate such as species (a) or (b) in Figure
4 as well as an increase in the rate of the transformation of
such a species to chloroacetylene. Second, raising the pH
could cause an increase in the rate of transformation of TCE
to chloroacetylene by a pathway involving a single elementary
reaction step, such as nucleophile-induced dichloroelimi-
nation (illustrated in Figure 4). Although it is not possible to
distinguish between these two mechanisms based on the
data reported here, both mechanisms can be explained by
a pH-dependent equilibrium between the acid and conjugate
base forms of FeS surface species, proposed previously (22)
to explain the pH-dependence of the rate of reductive
dechlorination of hexachloroethane (HCA) by FeS. Consid-
9
3 8 8 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

reduction potential of reactive mineral surface species. In
another study, Pecher et al. (74) studied the transformation
of CT and dibromodichloromethane by Fe(II)-coated oxide
surfaces and found that both disappearance rates and the
ratios of dihaloelimination to hydrogenolysis products
increased with higher concentrations of sorbed Fe(II). The
predominance of the dihaloelimination product (formate)
at high concentrations of sorbed Fe(II) was attributed to
differences in electron availability for the more Fe(II)-rich
surface (72, 74). This suggests that electron availability at the
mineral-water interface may influence the predominant
reaction pathway for TCE transformation by FeS and that
increased pH may affect rate constants and branching ratios
by increasing electron availability at the FeS surface.
Transform ation of TCE by FeSsInfluence of Solution
Am endm ents and FeS Aging. Several organic and inorganic
solutes were added to FeS aqueous slurries prior to or
concurrent with TCE addition in order to determine their
influence on reaction rates and pathways. Table 1 illustrates
that for a single pH value (8.3), addition of a variety of solution
amendments (MnCl2, 2,2′-bipyridine, and 2-propanol) did
not significantly affect pseudo-first-order rate constants or
branching ratios for TCE transformation by FeS.
MnCl2 was added to FeS slurries because Mn2+ has been
shown to adsorb to a significant extent to the FeS surface
(75), presumably to surface S atoms. Similarly, 2,2′-bipyridine
was added because it has a strong thermodynamic driving
force for complex formation with iron (76) and has been
shown to adsorb to a significant extent to the FeS surface
(22). Neither 1 mM Mn2+ nor 0.1 mM 2,2′-bipyridine, however,
affected the rate constants or product distributions for TCE
transformation by FeS. Calculation of approximate surface
Fe and S atom densities using the mackinawite unit cell
composition and dimensions given in refs 13 and 77 along
with the experimentally measured FeS specific surface area
of 0.05 m2/ g indicates that addition of 1 mM MnCl2 or 0.1
mM 2,2′-bipyridine to 10 g/ L FeS slurries would produce
Mn2+ or 2,2′-bipyridine at concentrations well in excess of
available surface Fe or S adsorption sites.
The result that 2,2′-bipyridine had no effect on the rate
of TCE transformation is distinct from that observed in the
transformation of HCA by FeS, where the presence of 2,2′-
bipyridine significantly increased the HCA transformation
rate. The lack of influence of MnCl2 and 2,2′-bipyridine on
rate constants or branching ratios suggests that a close
interaction between TCE and the FeS surface is not required
for electron transfer or else that surface sites for sorption of
Mn2+ and 2,2′-bipyridine are distinct from those where
electron transfer to TCE takes place.
2-Propanol was added as a solute to FeS slurries in order
to test whether, as a better hydrogen atom donor than
methanol, it would influence the distribution of reaction
products by favoring hydrogen atom abstraction by a
dichlorovinyl radical intermediate (Figure 4, species a) to
form cis-DCE versus transformation of such a radical
intermediate to a dichlorovinyl anion (Figure 4, species b)
or to chloroacetylene. The lack of influence of 2-propanol on
the TCE product distribution can be attributed to the
unimportance of hydrogen atom abstraction by a radical
intermediate such as species (a) or to the fact that neither
methanol nor 2-propanol are important hydrogen atom
donors in the aqueous FeS system.
FIGURE 5. Influence of NaHS on the transformation of TCE by 100
g/L Fisher iron at pH 8.3. The data points are experimentally measured
t
values, and the lines are plots of [TCE]aq,0e-k
.
obs
oxidation state of iron sulfide minerals in the environment
will strongly influence the potential for intrinsic remediation
of pollutants such as TCE. This is very relevant because
minerals such as greigite (Fe3S4) and pyrite (FeS2) have been
shown to form upon aging of FeS in sulfidic systems (6, 15,
30, 31). The oxidation of mackinawite to greigite is faster at
acidic versus alkaline pH values (15), suggesting that mineral
aging in low pH groundwaters may eventually result in
deactivation of iron sulfide mineral surfaces with respect to
reductive dechlorination. On the other hand, mackinawite
formed in higher pH environments, through reaction of iron
minerals or iron metal with sulfate reducing bacteria, may
be stable for longer periods. For example, mackinawite
formed through reaction of synthetic goethite (R-FeOOH)
with sulfate reducing bacteria at pH 8 was stable for at least
6-9 months (34).
Transform ation of TCE by Iron Metal Treated with NaHS.
Addition of 1 mM NaHS to both Fisher and Peerless iron
granules significantly increased the rate constants for TCE
transformation by these materials (Table 2). The transforma-
tion of TCE by Fisher iron granules, with and without the
addition of 1 mM NaHS, is illustrated in Figure 5. These results
are consistent with previous reports of increased rates of
TCE and PCE (27, 28) and CT (16, 29) degradation upon
addition of aqueous sulfide species to iron metal. This rate
increase has been attributed to the presence of dissolved
sulfide (16), the corrosion of iron metal by dissolved sulfide
(16), and the formation of FeS at the iron metal surface (16,
27, 28). This last explanation is supported by numerous
studies that have shown formation of iron sulfide, initially
in the form of mackinawite at neutral to alkaline pH values
(4, 7, 15), upon treatment of iron metal with aqueous sulfide
(3-14).
Shoesmith et al. (10) identified two forms of mackinawite
after treating polished iron samples with H2S-saturated
aqueous solutions at pH 7: a “coherent base layer” formed
by a solid state reaction, possibly beginning with chemi-
sorption of HS- to the iron surface, and a “loose deposit of
precipitated material at a few sites”. Based on this morpho-
logical description, at least two possible mechanisms for the
rate enhancement observed here are possible. First, a base
layer or film of mackinawite on the iron surface could act as
a conducting coating, increasing the rate of electron transfer
(73). Mackinawite possesses a layered structure with each
iron atom in square planar coordination with four other iron
atoms in each layer (13, 77). X-ray diffraction analysis
indicated that mackinawite layers in the “coherent base layer”
identified by Shoesmith et al. were oriented perpendicular
to the iron metal surface at pH 7, while at pH 6 and below,
both parallel and perpendicular orientations were found (10).
Unlike the previously discussed results, aged FeS was
completely unreactive with respect to TCE over 6.5 months,
yet this aged FeS transformed HCA to PCE with a rate constant
only slightly lower than that for nonaged FeS (data not
shown). This difference in reactivity could be due to the fact
that the reduction potential of the reactive species associated
with aged FeS was lower than that of HCA but higher than
that of TCE. This finding suggests that, not surprisingly, the
9
VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 8 8 9

(13) Lennie, A. R.; Redfern, S. A. T.; Schofield, P. F.; Vaughan, D. J.
Mineralogical Magazine 1995, 59, 677-683.
(14) Lennie, A. R.; Vaughan, D. J. In Mineral Spectroscopy: A Tribute
to Roger G. Burns; Dyar, M. D., McCammon, C., Schaefer, M.
W., Eds.; The Geochemical Society: Special Publication No. 5,
St. Louis, MO, 1996; pp 117-131.
(15) Rickard, D. T. Stockholm Contr. Geol. 1969, 26, 67-95.
(16) Lipczynska-Kochany, E.; Harms, S.; Milburn, R.; Sprah, G.;
Nadarajah, N. Chemosphere 1994, 29, 1477-1489.
(17) Sivavec, T. M.; Horney, D. P.; Baghel, S. S. Emerging Technologies
in Hazardous Waste Management VII; American Chemical
Society Special Symposium, September 17-20, 1995, Atlanta,
GA; American Chemical Society: Washington, DC, 1995; pp 42-
45.
Mackinawite is a metallic conductor (78) with the conductivity
or delocalized electrons in the plane of mackinawite layers
(8). Consequently, orientation of mackinawite layers per-
pendicular to the iron surface, as was observed at pH 7 (10),
could facilitate conduction of electrons from the iron metal
to an adsorbed compound such as TCE, resulting in increased
rates of TCE transformation. Second, the increased rate of
TCE transformation upon treatment of iron metal with NaHS
could be explained by reaction of precipitated mackinawite
deposits with TCE, since FeS is significantly more reactive
with TCE than iron metal on a surface area normalized basis
(21).
Unlike experiments with FeS alone, where cis-DCE and
acetylene were the sole reaction products detected in
significant quantities, significant quantities of 1,1-DCE,
ethylene, and ethane were also detected in the transformation
of TCE by sulfide-treated iron metal, precluding calculation
of branching ratios. The results illustrated in Figure 5,
however, indicate that treatment of iron metal with bisulfide
may hold promise as an effective means to increase the
reactivity of iron metal prior to or after installation in a
subsurface reactive barrier. The results shown in Figure 5
also suggest that encouraging the growth of sulfate-reducing
bacteria in the vicinity of zerovalent iron subsurface walls
may be an effective way to enhance the long-term reactive
stability of these walls through the formation of FeS surface
coatings or deposits on iron metal or through reactivation
of an aged and oxidized iron metal surface to form FeS.
(18) Sivavec, T. M. U.S. Patent Number 5,447,639, 1995.
(19) Sivavec, T. M.; Horney, D. P.; Baghel, S. S. U.S. Patent Number
5,575,927, 1996.
(20) Assaf, N.; Lin, K.-Y.; Manony, J. Preprints of Papers Presented at
the 213th ACS National Meeting, April 13-17, 1997, San
Francisco, CA; American Chemical Society: Washington, DC,
1997; Vol. 37 (1), pp 194-195.
(21) Sivavec, T. M.; Horney, D. P. Preprints of Papers Presented at
the 213th ACS National Meeting, April 13-17, 1997, San
Francisco, CA; American Chemical Society: Washington, DC,
1997; Vol. 37 (1), pp 115-117.
(22) Butler, E. C.; Hayes, K. F. Environ. Sci. Technol. 1998, 32, 1276-
1284.
(23) Butler, E. C.; Hayes, K. F. Environ. Sci. Technol. 1999, 33, 2021-
2027.
(24) Devlin, J. F.; Mu¨ ller, D. Environ. Sci. Technol. 1999, 33, 1021-
1027.
(25) Butler, E. C.; Hayes, K. F. Environ. Sci. Technol. 2000, 34, 422-
429.
(26) Kenneke, J. F.; Weber, E. J. Preprints of Papers Presented at the
220th ACS National Meeting, August 20-24, 2000, Washington,
DC; American Chemical Society: Washington, DC, 2000; Vol.
40 (2), pp 313-315.
(27) Hassan, S. M. Chemosphere 2000, 40, 1357-1363.
(28) Hassan, S. M.; Wolfe, N. L.; Cippolone, M. G. Preprints of Papers
Presented at the 209th ACS National Meeting, April 2-7, 1995,
Anaheim, CA; American Chemical Society: Washington, DC,
1995; Vol. 35 (1), pp 735-737.
(29) Harms, S.; Lipczynska-Kochany, E.; Milburn, R.; Sprah, G.;
Nadarajah, N. Preprints of Papers Presented at the 209th ACS
National Meeting, April 2-7, 1995, Anaheim, CA; American
Chemical Society: Washington, DC, 1995; Vol. 35 (1), pp 292-
294.
Acknowledgments
We thank Tom Yavaraski for invaluable technical assistance
in the laboratory and three anonymous reviewers for
comments that greatly improved the manuscript. Funding
was provided by the U.S. Environmental Protection Agency,
U.S. Department of Energy, National Science Foundation,
and Office of Naval Research Joint Program on Bioreme-
diation (EPA-G-R-825958). The content of this publication
does not necessarily reflect the views of these agencies.
Supporting Information Available
(30) Uda, M. American Mineralogist 1965, 50, 1487-1489.
(31) Sweeney, R. E.; Kaplan, I. R. Econ. Geol. 1973, 68, 618-634.
(32) Taylor, P.; Rummery, T. E.; Owen, D. G. J. Inorg. Nucl. Chem.
1979, 41, 596-599.
(33) Rickard, D. T. Am. J. Sci. 1975, 275, 636-652.
(34) Rickard, D. T. Stockholm Contr. Geol. 1969, 26, 49-66.
(35) Doyle, R. W. Am J. Sci. 1968, 266, 980-994.
(36) Emerson, S. Geochim. Cosmochim. Acta 1976, 40, 925-934.
(37) Davison, W. Aquatic Sciences 1991, 53, 309-321.
(38) Skinner, B. J.; Erd, R. C.; Grimaldi, F. S. American Mineralogist
1964, 49, 543-555.
(39) Campbell, T. J.; Burris, D. R.; Roberts, A. L.; Wells, J. R. Environ.
Toxicol. Chem. 1997, 16, 625-630.
(40) Arnold, W. A.; Roberts, A. L. Environ. Sci. Technol. 2000, 34,
1794-1805.
(41) Tezuka, M.; Yajima, T. Denki Kagaku Oyobi Kogyo Butsuri Kagaku
1991, 59, 517-518.
(42) Nagaoka, T.; Yamashita, J.; Kaneda, M.; Ogura, K. J. Electroanal.
Chem. 1992, 355, 187-195.
(43) Burris, D. R.; Delcomyn, C. A.; Smith, M. H.; Roberts, A. L.
Environ. Sci. Technol. 1996, 30, 3047-3052.
(44) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell,
T. J. Environ. Sci. Technol. 1996, 30, 2654-2659.
(45) Blo¨ chl, E.; Keller, M.; Wa¨chtersha¨user, G.; Stetter, K. O. Proc.
Natl. Acad. Sci. U.S.A. 1992, 89, 8117-8120.
(46) Osipov, A. M.; Boiko, Z. V.; Afanas’eva, L. N.; Grishchuk, S. V.
Solid Fuel Chemistry 1994, 28 (2), 9-12.
(47) Lee, W.; Batchelor, B. Preprints of Papers Presented at the 220th
ACS National Meeting, August 20-24, 2000, Washington, DC;
American Chemical Society: Washington, DC, 2000; Vol. 40 (2),
pp 338-340.
Methods used to calculate the 95% confidence intervals
reported in Tables 1 and 2. This material is available free of
charge via the Internet at http:/ / pubs.acs.org.
Literature Cited
(1) U.S. Environmental Protection Agency. Field Applications of In
Situ Remediation Technologies: Permeable Reactive Barriers;
EPA 542-R-99-002; Office of Solid Waste and Emergency
Response: Washington, DC, 1999.
(2) U.S. Environmental Protection Agency. Common Chemicals
found at Superfund Sites; EPA 540/ R-94/ 044; Office of Emer-
gency and Remedial Response: Washington, DC, 1994; as up-
dated at URL: http:/ / www.epa.gov/ superfund/ oerr/ comtools/
index/ htm#generalpublications, April 1998.
(3) Meyer, F. H.; Riggs, O. L.; McGlasson, R. L.; Sudbury, J. D.
Corrosion 1958, 14 (2), 69-75.
(4) Berner, R. A. Science 1962, 137, 669.
(5) Berner, R. A. J. Geol. 1964, 72, 293-306.
(6) Berner, R. A. Am. J. Sci. 1967, 265, 773-785.
(7) Takeno, S.; Zoˆ ka, H.; Nihara, T. Am. Mineral. 1970, 55, 1639-
1649.
(8) Vaughan, D. J.; Ridout, M. S. J. Inorg. Nucl. Chem. 1971, 33,
741-746.
(9) Pankow, J. F.; Morgan, J. J. Environ. Sci. Technol. 1979, 13, 1248-
1255.
(10) Shoesmith, D. W.; Taylor, P.; Bailey, G.; Owen, D. G. J.
Electrochem. Soc. 1980, 127, 1007-1015.
(11) Wikjord, A. G.; Rummery, T. E.; Doern, F. E.; Owen, D. G.
Corrosion Sci. 1980, 20, 651-671.
(12) Murowchick, J. B.; Barnes, H. L. Am. Mineralogist 1986, 71, 1243-
1246.
(48) Weerasooriya, R.; Dharmasena, B. Chemosphere 2001, 42, 389-
396.
(49) Sweeney, R. E.; Kaplan, I. R. Econ. Geol. 1973, 68, 618-634.
9
3 8 9 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

(50) Skinner, B. J.; Erd, R. C.; Grimaldi, F. S. American Mineralogist
1964, 49, 543-545.
(68) Ahr, H. J.; King, L. J.; Nastainczyk, W.; Ullrich, V. Biochem.
Pharmac. 1980, 29, 2855-2861.
(51) Good, N. E.; Winget, G. D.; Winter, W.; Connolly, T. N.; Izawa,
S.; Singh, R. M. M. Biochemistry 1966, 5, 467-477.
(52) Picardal, F. W.; Kim, S.; Radue, A.; Backhus, D. In Emerging
Technologies in Hazardous Waste Management 7; Tedder,
Pohland, Eds.; Plenum Press: New York, 1997; pp 81-90.
(53) Semadeni, M.; Chiu, P.-C.; Reinhard, M. Environ. Sci. Technol.
1998, 32, 1207-1213.
(54) Butler, E. C. Ph.D. Dissertation, University of Michigan, 1998.
(55) Leighton, D. T., Jr.; Calo, J. M. J. Chem. Eng. Data 1981, 26,
382-385.
(56) Patterson, J. W. Industrial Wastewater Treatment Technology,
2nd ed.; Butterworth: Boston, 1982; pp 320-322.
(57) Gossett, J. M. Environ. Sci. Technol. 1987, 21, 202-208.
(58) Munz, C.; Roberts, P. V. Am. Water Works Assn. J. 1987, 79(5),
62-69.
(59) Peng, J.; Wan, A. Environ. Sci. Technol. 1997, 31, 2988-3003.
(60) Handbook of Physical Properties of Organic Chemicals; Howard,
P. H., Meylan, W. M., Eds.; CRC: Lewis Publishers: Boca Raton,
1997.
(69) Krone, U. E.; Thauer, R. K.; Hogenkamp, H. P. C.; Steinbach, K.
Biochemistry 1991, 30, 2713-2719.
(70) Kriegman-King, M. R.; Reinhard, M. Environ. Sci. Technol. 1992,
26, 2198-2206.
(71) Kriegman-King, M. R.; Reinhard, M. Environ. Sci. Technol. 1994,
28, 692-700.
(72) Pecher, K.; Haderlein, S. B.; Schwarzenbach, R. P. Preprints of
Papers Presented at the 213th ACS National Meeting, April 13-
17, 1997, San Francisco, CA; American Chemical Society:
Washington, DC, 1997; Vol. 37 (1), pp 185-187.
(73) Balko, B. A.; Tratnyek, P. G. J. Phys. Chem. B 1998, 102, 1459-
1465.
(74) Pecher, K.; Kneedler, E. M.; Tonner, B. P. Preprints of Papers
Presented at the 217th ACS National Meeting, March 21-25,
1999, Anaheim, CA; American Chemical Society: Washington,
DC, 1999; Vol. 39 (1), pp 292-294.
(75) Arakaki, T.; Morse, J. W. Geochim. Cosmochim. Acta 1993, 57,
9-14.
(61) Steinfeld, J. I.; Francisco, J. S.; Hase, W. L. Chemical Kinetics
and Dynamics; Prentice Hall: Englewood Cliffs, NJ, 1989; pp
33-35.
(76) Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum
Press: New York, 1976.
(62) Bevington, P. R. Data Reduction and Error Analysis for the
Physical Sciences; McGraw-Hill: New York, 1993; Chapter 4.
(63) Glod, G.; Angst, W.; Holliger, C.; Schwarzenbach, R. P. Environ.
Sci. Technol. 1997, 31, 253-260.
(64) Seiber, J. N. J. Org. Chem. 1971, 36, 2000-2002.
(65) Fallab, S. Angew. Chem. Int. Ed. Engl. 1967, 6, 496-507.
(66) Conway, B. E.; Ku, J. C. H.; Ho, F. C. J. Colloid Interface Sci. 1980,
75, 357-372.
(77) Taylor, L. A.; Finger, L. W. Carnegie Institution of Washington
Year Book 1970, 69, 318-322.
(78) Welz, D.; Rosenberg, M. J. Phys. C: Solid State Phys. 1987, 20,
3911-3924.
Received for review February 7, 2001. Revised manuscript
received July 16, 2001. Accepted July 17, 2001.
(67) Wang, J. Analytical Electrochemistry; VCH: New York, 1994; p
2.
ES010620F
9
VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 8 9 1