Isotopic fractionation during reductive dechlorination of trichloroethene by zero-valent iron: Influence of surface treatment

Article abstract of DOI:10.1016/S0045-6535(02)00327-2

During reductive dechlorination of trichloroethene (TCE) by zero-valent iron, stable carbon isotopic values of residual TCE fractionate significantly and can be described by a Rayleigh model. This study investigated the effect of observed reaction rate, surface oxidation and iron type on isotopic fractionation of TCE during reductive dechlorination. Variation of observed reaction rate did not produce significant differences in isotopic fractionation in degradation experiments. However, a small influence on isotopic fractionation was observed for experiments using acid-cleaned electrolytic iron versus experiments using autoclaved electrolytic iron, acid-cleaned Peerless cast iron or autoclaved Peerless cast iron. A consistent isotopic enrichment factor of ε = -16.7‰ was determined for all experiments using cast iron, and for the experiments with autoclaved electrolytic iron. Column experiments using 100% cast iron and a 28% cast iron/72% aquifer matrix mixture also resulted in an enrichment factor of -16.9‰. The consistency in enrichment factors between batch and column systems suggests that isotopic trends observed in batch systems may be extrapolated to flowing systems such as field sites. The fact that significant isotopic fractionation was observed in all experiments implies that isotopic analysis can provide a direct qualitative indication of whether or not reductive dechlorination of TCE by Fe⁰ is occurring. This evidence may be useful in answering questions which arise at field sites, such as determining whether TCE observed down-gradient of an iron wall remediation scheme is the result of incomplete degradation within the wall, or of the dissolved TCE plume bypassing the wall.

Full text of DOI:10.1016/S0045-6535(02)00327-2

Chemosphere 49 (2002) 587–596
www.elsevier.com/locate/chemosphere
Isotopic fractionation during reductive dechlorination
of trichloroethene by zero-valent iron: influence
of surface treatment
a,b,
b
c
d
G.F. Slater
*, B. Sherwood Lollar , R. Allen King , S. O’Hannesin
a
Department of Marine Chemistry and Geochemistry, MS#4, Woods Hole Oceanographic Institution, 520 Woods Hole Road,
Woods Hole, MA 02543-1543, USA
b
Stable Isotope Laboratory, Department of Geology, University of Toronto, 22 Russell St., Toronto, Canada M5S 3B1
c
Department of Geology, Washington State University, P.O. Box 642812, Pullman, WA, USA
Envirometal Technologies, Inc., 745 Bridge St., Suite 7, Waterloo, Ontario, Canada N2V 2G6
d
Received 31 December 2001; received in revised form 10 June 2002; accepted 17 June 2002
Abstract
During reductive dechlorination of trichloroethene (TCE) by zero-valent iron, stable carbon isotopic values of re-
sidual TCE fractionate significantly and can be described by a Rayleigh model. This study investigated the effect of
observed reaction rate, surface oxidation and iron type on isotopic fractionation of TCE during reductive dechlori-
nation. Variation of observed reaction rate did not produce significant differences in isotopic fractionation in
degradation experiments. However, a small influence on isotopic fractionation was observed for experiments using acid-
cleaned electrolytic iron versus experiments using autoclaved electrolytic iron, acid-cleaned Peerless cast iron or
autoclaved Peerless cast iron. A consistent isotopic enrichment factor of e ¼ À16:7‰ was determined for all experi-
ments using cast iron, and for the experiments with autoclaved electrolytic iron. Column experiments using 100% cast
iron and a 28% cast iron/72% aquifer matrix mixture also resulted in an enrichment factor of À16.9‰. The consistency
in enrichment factors between batch and column systems suggests that isotopic trends observed in batch systems may be
extrapolated to flowing systems such as field sites.
The fact that significant isotopic fractionation was observed in all experiments implies that isotopic analysis can
provide a direct qualitative indication of whether or not reductive dechlorination of TCE by Fe0 is occurring. This
evidence may be useful in answering questions which arise at field sites, such as determining whether TCE observed
down-gradient of an iron wall remediation scheme is the result of incomplete degradation within the wall, or of the
dissolved TCE plume bypassing the wall.
Ó 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Dechlorination; Carbon isotopes; Iron; Acid-washed
1. Introduction
Reductive dechlorination of dissolved chlorinated
ethenes by in situ passive iron wall remediation schemes
* Corresponding author. Address: Department of Marine
has been of great interest in contaminant hydrogeology
Chemistry and Geochemistry, MS#4, Woods Hole Oceano-
since its initial suggestion by Gillham and O’Hannesin
graphic Institution, 520 Woods Hole Road, Woods Hole, MA
(1994). Research on this process has focused on deter-
mining the mechanisms and controls on reductive
02543-1543, USA. Fax: +1-508-457-2164.
E-mail address: gslater@whoi.edu (G.F. Slater).
0045-6535/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved.
PII: S0045-6535(02)00327-2

588
G.F. Slater et al. / Chemosphere 49 (2002) 587–596
dechlorination by iron metal, including research on
pathways (Matheson and Tratnyek, 1994; Orth and
Gillham, 1996; Roberts et al., 1996; Weber, 1996;
Charlet et al., 1998; Fennelly and Roberts, 1998; Arnold
and Roberts, 2000), kinetics (Johnson et al., 1996;
Charlet et al., 1998; Scherer et al., 1998), sorption effects
(Burris et al., 1998), hydrocarbon formation (Hardy and
Gillham, 1995; Deng et al., 1997), the role of surface
oxides (Sherer et al., 1997; Johnson et al., 1998; Su and
Puls, 1999), long term wall performance (O’Hannesin
and Gillham, 1998), and barrier design (Tratnyek et al.,
1997). Much progress has been made in understanding
the processes involved in reductive dechlorination by
iron metal which is presently thought to occur via
complexation of the chlorinated ethene to the metal
surface. Subsequent electron transfer from the metal to
the chlorinated ethene results in the loss of chlorine by
either b-elimination or hydrogenolysis (Arnold and
Roberts, 2000). However, questions remain concerning
the mechanisms of electron transfer and complexation,
and the role of oxides as barrier, semi-conductor or
coordinating surface (e.g. Sherer et al., 1997). While
there have been a number of successful field applications
of permeable reactive barriers (PRBs), several challenges
in monitoring the effectiveness of PRBs have arisen.
Monitoring iron wall performance can be complicated
by temporal variations in contaminant concentrations
(e.g. O’Hannesin and Gillham, 1998). Further, con-
firming the effectiveness of dechlorination in a PRBcan
be complicated by the observation of trichloroethene
(TCE) down-gradient from the wall. It is often impos-
sible to determine whether this TCE is present due to
bypass of the PRBthrough fracture flow or due to in-
complete reaction due to insufficient reactor thickness.
Stable carbon isotopic analysis is a potential tech-
nique for investigation and monitoring of reductive de-
chlorination of chlorinated ethenes by in situ granular
iron treatment systems. Stable carbon isotopic analysis
involves measurement of the ratio of the two stable
isotopes of carbon present in a sample. This ratio is
expressed as a d13C value where:
the residual chlorinated ethene pool can be enriched by
up to 40‰ during reductive dechlorination by zero-
valent iron. These shifts are highly reproducible and can
be modeled using a simple isotopic model known as the
Rayleigh model (Slater et al., 1998; Dayan et al., 1999).
The Rayleigh model assumes that there is a consistent
preferential reaction of one isotopomer over the other
during a process. This preference is expressed as a
fractionation factor, a, which relates the isotopic com-
position of the reactant to the extent of degradation as
in Eq. (2):
R=R0 ¼ f ðaÀ1Þ
ð2Þ
R is the ratio of 13C to 12C in the sample at a given
fraction remaining, f. R0 is the initial ratio of 13C to 12C.
This fractionation factor can be converted into a permil
enrichment factor, e, where e ¼ 1000ða À 1Þ.
This study characterized the isotopic effects associ-
ated with reductive dechlorination of TCE by both
electrolytic and cast iron, in order to assess whether
differences in iron type, surface pre-treatment or ob-
served reaction rate result in differences in isotopic
fractionation. Further, this study compared isotopic
fractionation in batch microcosm systems to column
systems containing 100% cast iron, and a mixture of cast
iron and aquifer material, in order to assess whether the
results observed in batch systems can be extrapolated
to open systems under dynamic flow conditions more
analogous to field systems. The long-term goal of the
research is to explore the utility of using stable carbon
isotopic analysis as a tool to investigate the mechanisms
and effectiveness of abiotic dechlorination in the labo-
ratory and the field.
2. Material and methods
This study consisted of three experiments using dif-
ferent types of iron and surface pre-treatments. The two
types of iron used were electrolytic iron and cast iron.
Electrolytic iron is essentially free from impurities and
is commonly used in laboratory studies of this reaction.
In contrast, cast iron contains impurities, in particular
graphite, and is used in field applications of this reme-
diation technology. These two iron types were compared
in order to determine whether the purity of the iron has
a significant effect on isotopic fractionation during re-
ductive dechlorination. The cast irons used in this study
were from two different suppliers, Peerless and Connelly.
These companies supply cast iron filings which are used
in field applications of this remediation technology.
Some differences in behavior have been observed during
the use of iron from these suppliers so the isotopic
fractionation by case iron from these two sources was
compared. Two different surface pre-treatments were
d13C ¼ ðRsample=Rstandard À 1Þ Â 1000
ð1Þ
Rsample is the 13C/12C ratio in a given sample, Rstandard is
the 13C/12C ratio in a standard reference material, in this
case V-PDB. Non-degradative processes such as vola-
tilization from groundwater and sorption do not result
in significant changes, or fractionation, of d13C, within
the 0.5‰ accuracy and reproducibility of online gas
chromatography-combustion-isotope ratio mass spectro-
metry (GC-C-IRMS) (Poulson and Drever, 1999; Slater
et al., 1999). In contrast, reductive dechlorination by
zero-valent iron results in a large stable carbon isotope
fractionation of chlorinated ethenes (Slater et al., 1998;
Dayan et al., 1999). The isotopic composition, d13C, of

G.F. Slater et al. / Chemosphere 49 (2002) 587–596
589
also compared in this study. Acid washing is commonly
used in laboratory studies to remove/reduce the oxidized
coating on the iron filings and thus increase the reac-
tivity of the iron. In the field no surface pre-treatment is
used and thus the surface is oxidized. Autoclaving was
used in these studies to provide sterile iron with an oxi-
dized surface which would be more analogous to that
used in the field.
vials containing water, pentane, TCE and acid washed
glass beads to make up the volume occupied by the iron.
Initial TCE concentrations were 20 mg/l.
2.2.1. Experiment 1
In order to compare isotopic fractionation at different
observed reaction rates, the iron/water ratio was varied
in batch systems using the same autoclaved Peerless cast
iron. The objective was to test the effect of variation in
observed reaction rate, rather than surface normalized
reaction rate, as this parameter may vary in the field. The
goal of the experiment was to determine whether such a
variation in the field would affect isotopic fractionation.
Autoclaved iron was used in order to ensure sterility in
the experiments. Isotopic behavior during these experi-
ments could therefore be assumed to be due to abiotic
reductive dechlorination alone and not to concurrent
biological reductive dechlorination which also has iso-
topic effects associated with it (Hunkeler et al., 1999;
Sherwood Lollar et al., 1999; Bloom et al., 2000; Slater
et al., 2001) and has been shown to occur in some abiotic
degradation systems (Novak et al., 1998; Gu et al., 1999).
Experiments were run using three iron/water ratios, 0.1,
0.05 and 0.025 g/ml. Based on previous experiments it
was determined that these ratios would degrade the TCE
over a period of one week, two weeks, and four weeks
respectively. For each of the three iron/water ratios, 12
reaction vials were set up. At each sampling point, two
vials from each of the three series were analyzed.
Three experiments were carried out in this study.
Experiment 1 compared the isotopic fractionation dur-
ing reductive dechlorination of TCE by Peerless cast
iron at three different observed reaction rates by varying
the iron/water ratio. Experiment 2 compared isotopic
fractionation by two different types of iron, electrolytic
iron powder and Peerless cast iron, which had each
undergone two different types of pre-treatment, acid
washing and autoclaving. This experiment assessed dif-
ferences in isotopic fractionation between the two types
of iron, as well as the impact of pre-treatment on iso-
topic fractionation during reductive dechlorination.
Experiment 3 assessed the utility of isotopic analysis as
an indicator of reaction efficiency in dynamic flow-
through column studies. Field conditions were simulated
as closely as possible by using aquifer material and
groundwater from a contaminated site. Two column
studies were carried out, the first with a column con-
taining 100% cast iron, and the second with a column
containing 28% cast iron and 72% aquifer material.
2.1. Experimental set up
2.1.1. Chemicals
2.2.2. Experiment 2
All solvents used were high purity and were obtained
from the following suppliers. TCE: ACP Chemical Inc.,
grade 99.5%. Pentane: EMS Scientific, spectrophoto-
metric grade. Methanol: ACP Chemicals Inc., ACS re-
agent grade. HCL: BDH, ACS reagent grade. All water
was deionized water (conductivity > 17 MX). Two dif-
ferent lots of Iron Electrolytic Powder, certified were
obtained from Fisher Scientific. Peerless and Connelly
cast iron filings were supplied by Envirometal Technol-
ogies Inc.
Experiment 2 compared isotopic fractionation be-
tween two lots of electrolytic iron (Fisher Scientific) and
one batch of cast iron (Peerless cast iron, obtained from
Envirometal Technologies Inc.). The two lots of elec-
trolytic iron were compared to ensure there was no
variation between lots from the manufacturer. All ex-
periments were carried out in duplicate reaction vials
that were sampled repeatedly over the course of degra-
dation. For both types of iron, experiments were run
using (i) iron that had been acid-cleaned and dried an-
aerobically, and (ii) iron which had been autoclaved and
dried aerobically. Autoclaving or acid cleaning the iron
ensured that no microbial activity occurred during de-
gradation and therefore that isotopic fractionation was
solely due to abiotic reductive dechlorination. During
acid cleaning, the iron was soaked in degassed 1 N HCl
for 30 min, then rinsed five times with degassed deion-
ized water, rinsed once with acetone and then dried
under UHP Ar and stored under N2 until use (Roberts,
pers. comm.). Experiments using acid-cleaned iron were
assembled under nitrogen atmosphere using degassed,
deionized water to minimize oxidation of the iron during
initial stages of the experiment. The autoclaved iron
was dried under high purity air and experiments using
2.2. Batch systems
Batch experiments were carried out in 60 ml serum
vials capped with Teflon faced septa and an iron to
water ratio of 0.25 g/ml for electrolytic iron and 0.10 g/
ml for Peerless cast iron. Different iron water ratios were
used for the two iron types in order to match observed
degradation rates as closely as possible. Batch vials
contained degassed, deionized water, 20 ml headspace,
iron, 80 lM pentane as an internal standard and TCE
added as a high concentration (ꢀ500 000 mg/l) TCE/
methanol stock solution. Vials were placed on a rotator
at 6 rpm and 22 Æ 1 °C. Controls consisted of serum

590
G.F. Slater et al. / Chemosphere 49 (2002) 587–596
autoclaved iron were assembled under aerobic condi-
tions, resulting in oxidation of the iron surface.
tector temperature was 300 °C. The carrier gas was he-
lium and makeup gas was 5% methane and 95% argon,
with a flow rate of 30 ml/min.
For analysis of concentrations of TCE and its
breakdown products (cDCE, tDCE, VC, ethene and
acetylene) in the batch experiments, 300 ll of headspace
were removed from the microcosm with a gas-tight sy-
ringe and injected into a Varian 3300 Gas Chromato-
graph equipped with a 30 m  0:53 mm GS-Q megabore
column and an FID detector. The temperature program
used was 70 °C hold 1 min, increase to 200 °C at 26 °C/
min then increase to 225 °C at 5 °C/min and hold 2 min.
External standards were prepared having the same
volume of headspace and liquid as the experimental
vials and covering the range of concentrations observed
in the experiments. Reproducibility on standard ana-
lyses was Æ5%. Reproducibility on replicate injections
was Æ5%.
2.3. Column systems
2.3.1. Experiment 3
Experiment 3 investigated the effect of dynamic flow
conditions on isotopic fractionation during reductive
dechlorination by untreated Connelly cast iron. Column
experiments were carried out using Envirometal Tech-
nologies Incorporated (ETI) test columns constructed to
test the application of iron wall remediation at an in-
dustrial facility in California. The column construction
and analytical methods are described in detail in Gill-
ham and O’Hannesin (1994).
Briefly, two columns were constructed of Plexiglas
tubes with a length of 100 cm and an internal diameter
of 3.8 cm. Seven sampling ports were positioned along
the length at distances of 5, 10, 20, 30, 40, 60 and 80 cm
from the inlet end the samples could be collected from
the influent and effluent solutions of the column as well.
One column was packed with 100% Connelly cast
iron with a surface area of 1.8 m2/g. To assure a homo-
geneous mixture, aliquots of iron material were dry-
packed vertically in lift sections within the column. The
second column was packed with 28% by weight of the
same Connelly cast iron, with the balance consisting of
aquifer material from the industrial facility in CA. The
column tests were performed at room temperature and
the iron was not pre-treated, as would be the case in field
applications.
2.5. Isotopic analysis
For isotopic analysis in the column experiments, a 10
ml water sample was collected from a sample port into a
15 ml serum bottle capped with a Teflon coated silicon
septa. The samples were returned to the University of
Toronto and analyzed within two days. For isotopic
analysis, 300–1000 ll of headspace were removed from
the experimental vial (Experiments 1 and 2) or the 15 ml
sampling vial (Experiment 3) using a gas-tight syringe
and injected on a gas chromatograph-combustion-iso-
tope ratio mass spectrometer (GC-C-IRMS) consisting
of a Varian 3400 GC equipped with a 30 m  0:25 mm
DB-624 (J&W Scientific) capillary column connected
via a combustion interface to a Finnigan Mat 252 iso-
tope ratio mass spectrometer. While internal reproduc-
ibility based on triplicate sample injections is <0.3‰
for this system, datapoints are assigned an accuracy
and reproducibility (error bars) of 0.5‰ to incorporate
not only variation due to reproducibility but variation
due to different split settings as per Dempster et al.
(1997).
An Ismatec IPN pump was used to feed the site water
spiked with 40 mg/l TCE from a collapsible Teflon bag
to the influent end of the columns. The column was
sampled daily over time until steady state concentration
profiles were achieved. In the bench-scale tests, steady
state was defined as the time when volatile organic
compound concentrations versus distance profiles do
not change significantly between sampling events.
2.4. Concentration analysis
For the column experiments, samples were with-
drawn from the column ports via syringe. Analysis of
TCE, and its breakdown products, cis- and trans-
dichloroethene (cDCE and tDCE), and vinyl chloride
(VC), involved pentane extraction of the water sample
followed by analysis on a Hewlett Packard 5890 Series II
gas chromatograph equipped with a Hewlett Packard
7673 autosampler. The chromatograph was equipped
with a 63Ni electron capture detector and DB-624
megabore capillary column (30 m  0:538 mm ID, film
thickness 3 lm). The gas chromatograph had an initial
temperature of 50 °C, with a temperature time program
of 15 °C/min to a final temperature of 150 °C. The de-
2.6. Iron characterization
In order to assess the impact of acid-washing and
autoclaving on the surface of the iron, the irons used in
Experiment 2 were examined under SEM and by XRF.
XRF analysis was carried out on a Phillips PW 2404
XRF. Iron fillings were attached to an oxygen free
graphite cup using two sided tape. The oxygen blank
was less than the 0.1% detection limit for oxygen. Sur-
face areas of the iron were determined by Micrometerics
(Norcross, GA) using BET N2 adsorption.

G.F. Slater et al. / Chemosphere 49 (2002) 587–596
591
3. Results
isotopic composition of the residual TCE was up to 40‰
more positive than the initial isotopic composition of
À30.5‰. Such isotopic enrichment of the residual TCE
is consistent with a process that degrades 12C–12C bonds
at a faster rate than 13C–12C bonds. No significant
change in isotopic composition was observed for the
control vials within 0.5‰.
Fig. 1 shows that the relationship of d13CTCE to frac-
tion of TCE remaining follows an exponential pattern.
This type of exponential relationship is characteristic of
processes that can be described by a Rayleigh isotopic
evolution model. The Rayleigh model can be applied to
an irreversible one step process if the fractionation fac-
tor, a, is constant throughout the reaction (Mariotti
et al., 1981). Because a are characteristic of the reaction,
and do not depend on the extent to which the reaction
proceeds, fractionation factors are an effective means
of describing and comparing the isotopic fractionation
occurring during different experiments. Fractionation
factors for the Rayleigh model curves in Fig. 1 were de-
termined by plotting the data from Experiment 1 using
the approach of Mariotti et al. (1981) shown in Eq. (3):
3.1. Experiment 1
TCE was degraded with autoclaved Peerless iron at
three different iron/water ratios. Degradation for the one
week, two week and four week degradations could be
described with observed first order rate constants of
0.0038, 0.0028, and 0.0021 (hÀ1) (Table 1). Surface
normalized rate constants (KSA) were not determined as
the iron used was from the same lot and thus the surface
normalized rate constants should be identical. Degra-
dation products were consistent with those reported in
the literature, including cDCE, tDCE, VC, ethene, and
acetylene, but were not quantified. No change in TCE
concentration was observed in the control vials within
Æ5%.
3.1.1. Isotopic results
For all three rates of degradation, the residual TCE
became progressively more enriched in the heavier iso-
tope (13C) as dechlorination proceeded (Fig. 1). The final
ða À 1Þ ln f ¼ lnðððd13CTCE=1000Þ þ 1Þ
Table 1
=ððd13CTCE =1000Þ þ 1ÞÞ
ð3Þ
First order rate constants and enrichment factors for reductive
dechlorination of TCE by Peerless cast iron at three iron/water
ratios, Experiment 1
0
where d13CTCE and d13CTCE are the isotopic composi-
tions of the TCE at a given time and the initial time re-
0
Iron/water
ratio
Rate
(hÀ1
r2
e
r2
95% CI
spectively. A least squares regression of a plot of ln f vs
)
lnðððd13CTCE=1000Þ þ 1Þ=ððd13CTCE =1000Þ þ 1ÞÞ yields a
0
0.1
0.05
0.025
0.0038
0.0028
0.0021
0.94
0.86
0.98
À15.8
À15.2
À16.8
0.84
0.88
0.95
Æ2.9
Æ3.0
Æ2.5
slope of a À 1. Fractionation factors were converted to
permil enrichment factors, e ¼ 1000ða À 1Þ for compar-
ison.
Enrichment factors (e) determined for the three iron/
water ratios used in this experiment are shown in Table 1
along with their r2 correlation factors and 95% confi-
dence intervals for the Mariotti type least squares re-
gression used to determine the enrichment factors. The
high r2 factors obtained for the Mariotti type least
square regressions demonstrate that the isotopic be-
havior in these systems is consistent with a Rayleigh
model. The fact that the enrichment factors are the same
for all three rates of degradation shows that isotopic
fractionation is independent of rate over this range of
rates. When all the data for these three iron/water ratios
is combined the overall enrichment factor is À15.7‰,
with an r2 of 0.87 and a 95% confidence interval of Æ1.6.
This overall enrichment factor is the same as the indi-
vidual enrichment factors for each experiment within
95% confidence and hence in Fig. 1 a single Rayleigh
curve based on the overall value is used for all the data.
Early in the experiment (f > 0:4) there were some
datapoints which did not fall on the e ¼ À15:7 Rayleigh
curve. While this deviation could be the result of non-
equilibrium partitioning of TCE between the solid, liquid
Fig. 1. Carbon isotopic composition of TCE versus fraction of
TCE remaining during reductive dechlorination by Peerless
zero-valent cast iron at three different iron/water ratios in Ex-
periment 1. Error bars represent 0.5‰ accuracy and repro-
ducibility and are smaller than plot symbols. The line represents
a Rayleigh model curve based on an enrichment factor of
À15.7‰ determined from all data using Eq. (3) (see text).

592
G.F. Slater et al. / Chemosphere 49 (2002) 587–596
and gaseous phases in early stages of the experiment,
equilibration should have occurred by the time these
samples were taken. It is more likely that some variation
in the individual reaction vials sampled at these points,
such as a slight leak, resulted in deviation observed by
these samples. By repeatedly sampling the same reaction
vial in subsequent experiments this problem was avoi-
ded. This trend does not appear in results from Experi-
ment 2. It is important to note that removing this early
data results in an enrichment factor of À16.4‰ (r2 ¼
0:92, 95% CI Æ1.4), which is not significantly different
from the value obtained using all the data.
rate constant for the autoclaved Peerless iron was nearly
an order of magnitude lower than that observed for the
acid cleaned Peerless iron. This difference is the same as
that found for the observed rate. However, accurate
surface normalized rate constants, which take into ac-
count the presence of non-reactive graphite in the cast
iron, could not be calculated as the carbon content of
this Peerless iron was not known.
3.2.1. Isotopic results
Fig. 2 shows the isotopic composition of TCE versus
fraction remaining for acid-cleaned batch 1 electrolytic
iron (A), autoclaved batch 1 electrolytic iron (B), acid-
cleaned Peerless iron (C), and autoclaved Peerless iron
(D). The results of the batch 2 electrolytic iron are
similar to batch 1 and are therefore not shown in this
figure. Enrichment factors for batch 1 and batch 2
electrolytic iron, and for Peerless cast iron are given in
Table 2. The r2 correlation co-efficients and 95% confi-
dence intervals for the Mariotti type least squares re-
gression are also given in Table 2. The high r2 values
(0.92–0.99) demonstrate that the isotopic behavior of
these systems is consistent with a Rayleigh model.
Table 2 shows that enrichment factors for acid-
cleaned and autoclaved Peerless iron are the same within
95% confidence intervals. In addition, enrichment fac-
tors for both batches of autoclaved electrolytic iron are
the same within 95% confidence intervals as those ob-
served for Peerless cast iron. Only acid-cleaned electro-
lytic iron has enrichment factors that are significantly
different. The enrichment factors for the acid-cleaned
electrolytic iron in batch 1 and batch 2 are À20.3‰ and
À24.8‰, significantly more negative than enrichment
factors for autoclaved electrolytic iron with respect to
the error of the measurements.
3.2. Experiment 2
Degradation of TCE to less chlorinated ethenes and
ethene and acetylene was observed consistent with re-
ports in the literature, but these breakdown product
concentrations were not quantified. No change in TCE
concentration was observed in the control vials within
Æ5%. First order degradation rate constants also de-
scribe the data obtained in Experiment 2 (Table 2). Ob-
served reaction rates varied by over two orders of
magnitude between experiments. The highest observed
rate was for autoclaved electrolytic iron from batch 1
which had the highest surface area (0.2410 m2/g). The
surface area analysis of batch 1 acid-cleaned electrolytic
iron was approximately a factor of four lower (0.0569
m2/g). Surface normalized rate constants (KSA À1 mÀ2 l)
h
for these two surface treatments were very similar
(0.0081, 0.0071). This suggests that observed rate of re-
action for these two experiments is controlled by the
surface area of the iron, consistent with the literature
(Johnson et al., 1996). The surface areas of batch 2
electrolytic iron were not determined so no comparison
between the surface areas of these batches could be made.
There was less difference between surface areas of the
acid-cleaned and autoclaved Peerless irons (0.3757 and
0.3525 m2/g respectively). The highest observed rate did
correspond to the higher surface area however, consis-
tent with the electrolytic results. The surface normalized
3.3. Experiment 3
3.3.1. Isotopic results
Fig. 3 shows the results of the column studies. Both
columns showed dechlorination of >95% of the initial
Table 2
First order rate constants and enrichment factors for dechlorination of TCE by different zero-valent iron types and surface treatments,
Experiment 2
Iron type
Surface treatment
Rate (hÀ1
)
r2
KSA (hÀ1 mÀ2 l)
e
r2
95% CI
Electrolytic batch 1
Acid cleaned
Autoclaved
0.12
0.45
0.82
0.99
0.0081
0.0071
À20.3
À17.2
0.98
0.99
Æ1.7
Æ0.4
Electrolytic batch 2
Peerless
Acid cleaned
Autoclaved
0.11
0.0028
0.95
0.99
Nd
Nd
À24.8
À18.7
0.96
0.96
Æ1.9
Æ4.1ꢁ
Acid cleaned
Autoclaved
0.0091
0.0012
0.96
0.96
0.0027
0.00039
À17.1
0.98
0.92
Æ1.4
Æ1.7
À16.4
Nd: surface areas for batch 2 electrolytic iron were not determined.
* 95% confidence intervals were large due to the small number of datapoints obtained in this experiment.

G.F. Slater et al. / Chemosphere 49 (2002) 587–596
593
Fig. 2. A–D: Carbon isotopic composition of TCE versus fraction of TCE remaining during reductive dechlorination by zero-valent
iron in Experiment 2. A: Batch 1 Electrolytic iron-acid washed. B: Batch 1 Electrolytic-iron autoclaved. C: Peerless iron-acid washed.
D: Peerless iron-autoclaved. Error bars represent 0.5‰ accuracy and reproducibility and are smaller than plot symbols. Lines represent
Rayleigh model curves based on enrichment factors determined from all data for each experiment using Eq. (3) (see text).
TCE to less chlorinated products (cDCE, tDCE, VC)
before the end of the column. Isotopic enrichment of
TCE was observed in both the column experiments and
could be described by a Rayleigh model. The enrichment
factors for the two columns were À16.9‰ (r2 ¼ 0:89)
and À17.4‰ (r2 ¼ 0:99) for the 100% and 22% iron
columns respectively. The outlier point on Fig. 3 is likely
due to an error in the measurement of the concentration
at this point. The variation in this point from the rest of
the data resulted in large 95% confidence intervals. The
overall enrichment factor for the combined data from
both column experiments was À16.9‰ with an r2 of 0.90
and a 95% confidence interval of Æ4.1‰.
by zero-valent iron. This enrichment is consistent with
a reaction that degrades 12C containing molecules at a
slightly faster rate than 13C containing molecules. Such a
process can be modeled using the Rayleigh model
(Mariotti et al., 1981).
The results of Experiment 1 illustrated that enrich-
ment factors were independent of observed reaction rate
over a range of a factor of 2. The results obtained from
Experiment 2 confirm that there is no systematic effect of
rate on enrichment factor. In fact, there is a variation of 2
orders of magnitude in observed reaction rate between
autoclaved electrolytic experiments, yet there is only
slight (1.5‰) variation in enrichment factor. This sug-
gests that observed rate of reaction does not play a major
role in determining the isotopic fractionation during re-
ductive dechlorination by zero-valent iron over the range
of rates investigated in this study. This observation is im-
portant for the application of isotopic analysis to moni-
toring iron wall reductive dechlorination at field sites.
4. Discussion
In all experiments, significant isotopic enrichment of
the TCE was observed during reductive dechlorination

594
G.F. Slater et al. / Chemosphere 49 (2002) 587–596
surface morphology of the autoclaved versus acid-
cleaned electrolytic iron. Nor is there any systematic
relationship between enrichment factor and differences
in surface area or KSA between autoclaved and acid-
cleaned electrolytic iron.
This increased concentration of oxygen on the sur-
face of the autoclaved electrolytic iron corresponded to
a reduction in enrichment factor for the autoclaved
electrolytic iron relative to the acid-cleaned electrolytic
iron. The mechanism causing this relationship is not
clear. However, it is possible that oxidation may be af-
fecting the activation energy of the reaction. Iron with
surface oxidation and/or impurities may require higher
activation energy to transfer electrons to the TCE. Su
and Puls (1999) calculated that untreated iron (more
oxidized) had higher activation energies than acid-
cleaned iron (less oxidized). In addition, they point out
that impurities, different reactive sites on the metal
surface, structure and morphology, could all be playing
a role in affecting the activation energy of the reaction.
Further research is planned to determine whether the
effect of oxidation on the enrichment factor of this re-
action is due to a change in activation energy, or to some
other as yet unknown factor.
Fig. 3. Carbon isotopic composition of TCE versus fraction of
TCE remaining during reductive dechlorination by Connelly
zero-valent cast iron in column experiments in Experiment 3.
Error bars represent 0.5‰ accuracy and reproducibility and are
smaller than plot symbols. The line represents a Rayleigh model
curve based on an enrichment factor of À16.9‰ determined
from all data using Eq. (3) (see text).
4.1. Effect of surface treatment
While enrichment factors for the autoclaved versus
acid-cleaned Peerless cast iron are the same within 95%
confidence intervals (À16.4‰ and À17.1‰ respectively),
there is a slight absolute reduction in enrichment factor
for the autoclaved cast iron. Possibly for cast iron, the
isotopic effects due to oxidation during autoclaving are
not as significant relative to other factors, such as the
presence of carbon impurities.
Since the data from these experiments can be de-
scribed by a Rayleigh model, the enrichment factor from
each experiment can be used to compare isotopic frac-
tionation between experiments. Surface pre-treatment
had a significant effect on isotopic fractionation by
electrolytic iron. The acid-cleaned electrolytic irons had
enrichment factors of À20.3‰ and À24.8‰, while the
autoclaved electrolytic irons had enrichment factors of
À17.2‰ and À18.7‰ (Table 2). In contrast, there was
no significant effect on enrichment factor due to surface
pre-treatment of the Peerless cast iron. The enrichment
factors for acid-cleaned and autoclaved Peerless iron
were the same within 95% confidence intervals.
4.2. Implications for reaction mechanism studies
The relatively large e values reported in Table 1
(e ¼ À15 to À25) are consistent with a primary kinetic
isotope effect. Arnold and Roberts (2000) hypothesized
that formation of a di-r-bonded intermediate is the
rate determining step during abiotic reductive dechlori-
nation of chlorinated ethenes. They noted that the pri-
mary kinetic isotope effect observed (e ¼ À17 to À22)
for other reactions involving conversion of a p bond to a
di-r-bond are very similar to those reported in Table 1
(e ¼ À15 to À25). Thus, this isotopic evidence provides
support for their model of the abiotic reductive de-
chlorination mechanism.
The change in fractionation factor observed for
electrolytic iron is co-incident with oxidation of the iron
surface. XRF analysis indicates that autoclaving the
iron significantly increased the amount of oxygen pre-
sent on the iron surface. XRF analysis of the electrolytic
iron from Experiment 1 revealed that there were signif-
icant amounts of oxygen (1.5%) within the first 0.8 lm
of the autoclaved electrolytic iron, while there were in-
significant amounts of oxygen (<0.1%) within the first
0.8 lm of the acid-cleaned electrolytic iron. Because
XRF response decays exponentially, if the oxygen on the
autoclaved iron was evenly distributed, 50% of the oxy-
gen would be within 0.2 lm of the surface. However, in
this case it is likely that the oxygen was located on the
iron surface, and that the actual percentage of oxygen
on the surface was much higher than 1.5%. SEM ex-
amination did not show any observable effect on the
A comparison of this study with other fractionation
values reported in the literature shows that in some cases
there are variations in e which are larger than those
observed in this study. In some cases, e reported are the
same as this study, for instance, Bill et al. (2001), ob-
served an enrichment factor of À16‰ for dechlorination
€
of TCE by cast iron. In contrast, a study by Schuth et al.
(submitted for publication) reports an enrichment factor
of À10.1‰ for reductive dechlorination of TCE by cast

G.F. Slater et al. / Chemosphere 49 (2002) 587–596
595
iron removed from PRBs in Belfast, Ireland and Tueb-
ingen, Germany. In addition, Dayan et al. (1999) re-
ported a fractionation factor of À8.6‰ during reductive
dechlorination by acid-cleaned electrolytic iron, a much
lower value than observed in our study (À20.4‰,
À24.8‰). The reason for the observation of these lower
fractionation factors is not clear. The observed varia-
tions in isotopic fractionation do not appear to be re-
lated to observed rate of reaction, surface areas or
surface treatments. Iron from different sources was used
in each of these experiments. Dayan et al. (1999) used
similar to TCE up-gradient of the wall. If TCE observed
down-gradient from a PRBis determined to be due to
insufficient reaction time within the wall, extensive ac-
tion must be taken to increase the wall thickness or in-
stall a second wall. If the TCE is due to wall bypass, then
localization of the bypass zone and repair may be car-
ried out. While considerably more research must be
undertaken to constrain the precise controls on isotopic
fractionation during reductive dechlorination of TCE,
the consistent large isotopic enrichment in residual TCE
produced by abiotic degradation already has the po-
tential to be an effective tool in evaluating and opti-
mizing passive remediation barrier technology.
€
acid cleaned electrolytic iron, Schuth et al. used un-
treated cast iron and iron from a field installation. Bill
et al. (2001) used a cast iron similar to that used by
€
Schuth et al. Potentially the observed variations are due
to some difference in surface condition of the iron which
is not resolved by present descriptions of the surface pre-
treatment. Alternatively they may indicate that in some
cases conversion to a di-r-bonded intermediate is not
the rate determining step, or that some other as yet
unclear factor is involved in the expression of isotopic
fractionation during this process. Further study of the
causes of these variations in isotopic enrichment factors
may yield insights into the mechanisms of this reaction.
In addition, understanding of the controls on isotopic
fractionation during reductive dechlorination by zero-
valent iron is required before isotopic enrichment factors
can be used as a quantitative tool to monitor the reac-
tion progress and thus the performance in the complex
environment of an iron reactive barrier as suggested by
Acknowledgements
The authors wish to thank the organizations which
contributed funding for this project––the Natural Sci-
ences and Engineering Research Council of Canada
Strategic Grants program and the University Consor-
tium on Solvents-in-Groundwater. Thanks are also due
to N. Arner and H. Li for technical support.
References
Arnold, W.A., Roberts, A.L., 2000. Pathways and kinetics of
chlorinated ethylene and chlorinated acetylene reaction with
Fe(0) particles. Environ. Sci. Technol. 34, 1794–1805.
Bill, M., Schuth, C., Barth, J.A.C., Kalin, R., 2001. Carbon
isotope fractionation during abiotic reductive dehalogen-
ation of trichloroethylene (TCE). Chemosphere 44, 1281–
1286.
€
Schuth et al. (submitted for publication).
4.3. Field implications
Bloom, Y., Aravena, R., Hunkeler, D., Edwards, E., Frape,
S.K., 2000. Carbon isotope fractionation during microbial
dechlorination of trichloroethene, cis-1,2-dichloroethene,
and vinyl chloride: Implications for assessment of natural
attenuation. Environ. Sci. Technol. 34, 2768–2772.
Burris, D.R., Allen-King, R.M., Manoranjan, V.S., Campbell,
T.J., Loraine, G.A., Deng, B., 1998. Chlorinated ethene
reduction by cast iron: Sorption and mass transfer. J.
Environ. Eng. 124, 1012–1019.
Though isotopic enrichment factors cannot be ap-
plied quantitatively in the field at this time, the obser-
vation that reductive dechlorination by zero-valent iron
consistently produces a large isotopic enrichment in the
d13C value of the residual TCE indicates that isotopic
analysis can provide a qualitative indicator of the ef-
fectiveness of this remediation approach in the field.
Further, the consistency between the enrichment factors
for the batch and column experiments in this study in-
dicate that the large isotopic fractionations observed in
the batch studies should be observed in the field and
that the presence of other matricies, such as aquifer
material, will not affect the patterns of isotopic frac-
tionation. Potential applications of isotopic analysis to
field situations include determining whether TCE ob-
served down-gradient of an iron wall is the result of
insufficient reaction time within the wall, or due to the
groundwater plume bypassing the wall. Partially de-
graded TCE will have isotopically enriched d13C values
relative to TCE up-gradient of the iron wall, while TCE
which has bypassed the wall should have d13C values
Charlet, L., Liger, E., Gerasimo, P., 1998. Decontamination of
TCE- and U-rich waters by granular iron: Role of sorbed
Fe(II). J. Environ. Eng., 25–30.
Dayan, H., Abrajano, T., Sturchio, N.C., Winsor, L., 1999.
Carbon isotopic fractionation during reductive dehalogen-
ation of chlorinated ethenes by metallic iron. Org. Geo-
chem. 30, 755–763.
Dempster, H.S., Sherwood Lollar, B., Feenstra, S., 1997.
Tracing organic contaminants in groundwater: A new
methodology using compound-specific isotopic analysis.
Environ. Sci. Technol. 31, 3193–3197.
Deng, B., Campbell, T.J., Burris, D.R., 1997. Hydrocarbon
formation in metallic iron/water systems. Environ. Sci.
Technol. 31, 1185–1190.

596
G.F. Slater et al. / Chemosphere 49 (2002) 587–596
Fennelly, J., Roberts, L., 1998. Reaction of 1,1,1-trichloro-
ethane with zero-valent metals and bimetallic reductants.
Environ. Sci. Technol. 32, 1980–1988.
Roberts, A.L., Totten, L.A., Arnold, W.A., Burris, D.R.,
Campbell, T.J., 1996. Reductive elimination of chlorinated
ethylenes by zero-valent metals. Environ. Sci. Technol. 30,
2654–2659.
Gillham, R.W., O’Hannesin, S.F., 1994. Enhanced degradation
of halogenated aliphatics by zero-valent iron. Ground
Water 32, 958–967.
Roberts, A.L. Personal communication.
Scherer, M.M., Balko, B.A., Gallagher, D.A., Tratnyek, P.G.,
1998. Correlation analysis of rate constants for dechlorina-
tion by zero-valent iron. Environ. Sci. Technol. 32, 3026
(abstract).
Gu, B., Phelps, T.J., Liang, L., Dickey, M.J., Roh, Y., Kinsall,
B.L., Palumbo, A.V., Jacobs, G.K., 1999. Biogeochemical
dynamics in zero-valent iron columns: Implications for
permeable reactive barriers. Environ. Sci. Technol. 33,
2170–2177.
€
Schuth, C., Bill, M., Barth, J., Slater, G.F., Kalin, R.M. Long
term performance of iron reactive barriers––rates and
carbon isotope fractionation in batch-experiments with
original samples. J. Cont. Hydro., submitted for publica-
tion.
Hardy, L.I., Gillham, R.W., 1995. Formation of hydrocarbons
from the reduction of aqueous CO2 by zero-valent iron.
Environ. Sci. Technol. 30, 57–65.
Hunkeler, D., Aravena, R., Butler, B.J., 1999. Monitoring
microbial dechlorination of tetrachloroethene (PCE) in
groundwater using compound-specific stable carbon isotope
ratios: Microcosm and field studies. Environ. Sci. Technol.
33, 2733–2738.
Sherer, M.M., Westall, J.C., Ziomek-Moroz, M., Tratnyek,
P.G., 1997. Kinetics of carbon tetrachloride reduction at an
oxide-free iron electrode. Environ. Sci. Technol. 31, 2385–
2391.
Sherwood Lollar, B., Slater, G., Ahad, J., Sleep, B., Spivack, J.,
Brennan, M., MacKenzie, P., 1999. Contrasting carbon
isotope fractionation during biodegradation of trichloro-
ethylene and toluene: Implications for intrinsic bioremedi-
ation. Org. Geochem. 30, 813–820.
Johnson, T.L., Scherer, M.M., Tratnyek, P.G., 1996. Kinetics
of halogenated organic compound degradation by iron
metal. Environ. Sci. Technol. 30, 2634–2640.
Johnson, T.L., Fish, W., Gorby, Y.A., Tratnyek, P.G., 1998. De-
gradation of carbon tetrachloride by iron metal: Complexation
effects on the oxide surface. J. Contam. Hydrol. 29, 379–398.
Mariotti, A., Germon, J.C., Hubert, P., Kaiser, P., Letolle, R.,
Tardieux, A., Tardieux, P., 1981. Experimental determina-
tion of nitrogen kinetic isotope fractionation: Some princi-
ples; illustration for the denitrification and nitrification
processes. Plant Soil 62, 413–430.
Slater, G.F., Dempster, H.D., Sherwood Lollar, B., Spivack, J.,
Brennan, M., MacKenzie, P., 1998. Isotopic tracers of
degradation of dissolved chlorinated solvents. First Inter-
national Battelle Conference on Remediation of Chlori-
nated and Recalcitrant Compounds, Monterey, CA.
Slater, G.F., Dempster, H.S., Sherwood Lollar, B., Ahad, J.,
1999. Headspace analysis: A new application for isotopic
characterization of dissolved organic compounds. Environ.
Sci. Technol. 33, 190–194.
Matheson, L.J., Tratnyek, P.G., 1994. Reductive dehalogen-
ation of chlorinated methanes by iron metal. Environ. Sci.
Technol. 28, 2045–2053.
Slater, G.F., Sherwood Lollar, B., Edwards, E., Sleep, B., 2001.
Variability in carbon isotopic fractionation during biode-
gradation of chlorinated ethenes: Implications for field
applications. Environ. Sci. Technol. 35, 901–907.
Su, C., Puls, R.W., 1999. Kinetics of trichloroethene reduction
by zero-valent iron and tin: Pretreatment effect, apparent
activation energy, and intermediate products. Environ. Sci.
Technol. 33, 163–168.
Novak, P.J., Daniels, L., Parkin, G.F., 1998. Enhanced
dechlorination of carbon tetrachloride and chloroform in
the presence of elemental iron and methanosarcina barkeri,
methanosarcina thermophila, or methanosaeta cobcillii.
Environ. Sci. Technol. 32, 1438–1443.
O’Hannesin, S.F., Gillham, R.W., 1998. Long-term prefor-
mance of an in situ ‘‘iron wall’’ for remediation of VOCs.
Ground Water 36, 164–170.
Tratnyek, P.G., Johnson, T.L., Scherer, M.M., Eykholt, G.R.,
1997. Remediating ground water with zero-valent metals;
chemical considerations in barrier design. Ground Water
Monit. Remed. XVII, 108–114.
Orth, S., Gillham, R.W., 1996. Dehalogenation of trichloro-
ethane in the presence of zero-valent iron. Environ. Sci.
Technol. 30, 66–71.
Poulson, S.R., Drever, J.I., 1999. Stable isotope (C, Cl, and H)
fractionation during vaporization of trichloroethylene. En-
viron. Sci. Technol. 33, 3689–3694.
Weber, E.J., 1996. Iron-mediated reductive transformations:
Investigation of reaction mechanism. Environ. Sci. Technol.
30, 716–719.