with no known chlorinated contaminants; no additional
purification was performed.
were incubated in the dark on a swirling shaker at 30 °C, and
samples were taken in duplicate and averaged.
Aqueous 2,3,5-TCB Am pule Experim ent. An ampule
experiment was conducted with aqueous 2,3,5-TCB to
observe the product distribution for an individual step of the
2,3,4,5,6-PeCB reductive dechlorination pathway. Ampules
were prepared with 0.5 mmol/ L vitamin B12, 7 mmol/ L
titanium citrate, and excess 2,3,5-TCB (200 µg/ mL nominal).
Ampules were incubated in the dark at 20 °C.
TCR Experim ents. Two-chambered reactors (TCRs) al-
lowed measurement of reductive dechlorination product
distributions over short time periods. TCRs were created as
described by Smith and Woods (23). The upper chamber of
the reactor was continuously purged, and reductive dechlo-
rination reactions continued for periods up to 8 days.
TCR experiments were conducted for 2,3,4,5-TeCB, 2,3,4,6-
TeCB, 2,3,5,6-TeCB, 2,3,4-TCB, 2,3,6-TCB, 2,4,5-TCB, 2,4,6-
TCB, and 3,4,5-TCB. For each PCB, 3 mL of a solution of 75
µg/ mL PCB in 0.5 mmol/ L B12 was introduced to a TCR. The
solution was purged with purified argon or nitrogen at 15
mL/ min for 10 min, and 30 µL of titanium(III) citrate was
dispensed into the reactor. TCRs were held at 20 °C using a
circulating water bath.
In previous experiments, the presence of the three forms
of vitamin B12 at various redox potentials was verified based
on spectrophotometric comparison with published spectra.
These studies confirmed that vitamin B12s was the dominant
form present in a solution containing excess titanium citrate
(21). In the current work, the form of vitamin B12 was verified
visually by observing the color change from red (B12a) to amber
(B12r) to dark blue (B12s).
Am pule Experim ents. Long-term experiments were con-
ducted in hermetically sealed glass ampules to demonstrate
reductive dechlorination of PCBs by vitamin B12s. Vitamin
B12 and PCBs in water/ methanol or sediment slurry were
dispensed into 2 mL borosilicate glass ampules. The solution
was purged with oxygen-scavenged nitrogen or argon (7-10
min at 15 mL/ min). Titanium(III) citrate was added to the
ampule using a syringe fitted with a 0.5 mm fused silica
capillary needle, and the ampule was quickly flame sealed.
In fully reduced ampules, the solution changed from bright
red (vitamin B12a) to amber (B12r) to dark blue (B12s) within
a few minutes. Improperly sealed ampules changed back to
amber or red and were discarded.
TCRs were sampled periodically to measure product
distributions for each reaction. Approximately 120 µL of
solution was drawn through the vent capillary into a small
culture tube. From this sample, 100 µL was delivered into a
2.7 mL glass vial with a Teflon screw cap. One milliliter of
hexane was added, and the vial was wrist shaken for 5 min.
The hexane fraction was concentrated into a 150 µL conical
autosampler vial to allow the detection of minor transfor-
mation products and a comparison of their relative con-
centrations.
Analytical Procedures. PCBs were analyzed by 1.0 µL
splitless injection with a Hewlett-Packard 7673 autosampler
coupled to an HP 5890 or 6890 gas chromatograph with
electron capture detector. Individual congeners were identi-
fied by retention time comparison with pure standards. The
column was a 30 m × 0.32 mm i.d. fused silica capillary
column with a 0.25 µm thickness DB-5 liquid phase (J & W
Scientific, Inc., Folsom, CA). Injector and detector temper-
atures were 300 °C. The oven temperature was programmed
as follows: 45 °C for 2 min, 25 °C/ min to 120 °C, 3.0 °C/ min
to 200 °C, 10 °C/ min to 245°, hold for 10 min. The carrier gas
was helium (13 psig head pressure), and the makeup gas was
95% argon/ 5% methane. The average error in our analysis
was (13% (SD/ mean), based on recovery of the internal
standard in the sediment ampule experiment.
Ampules were periodically sacrificed for sampling. The
ampule was broken at the neck, and 1 mL of hexane was
added directly to the ampule. The ampule was resealed with
a Teflon/ silicon cap and shaken on a wrist shaker for a
specified time (5 min in the aqueous experiments, 30 min
in the sediment experiment). The hexane fraction was
transferred to a 2 mL amber vial with a 300 µL glass insert
and black viton septum for analysis. These experiments were
conducted with PCB concentrations well above solubility
limits; the addition of hexane directly to the ampule reduced
error due to sorption/ desorption phenomena. For the
sediment ampule experiment, the hexane contained 10 µg/ L
2,3′,4,4′-TeCB as an internal standard to monitor the repeat-
ability of the extraction procedure. The reaction conditions
varied for each ampule experiment and are detailed below.
Aqueous 2,3,4,5,6-PeCB Am pule Experim ent. For the
aqueous ampule experiment, a total of 40 replicate ampules
was created at three separate times, and the results were
combined to record the progress curve for reductive dechlo-
rination of aqueous 2,3,4,5,6-PeCB by vitamin B12s. Each
ampule was prepared with 1 mL of a solution containing 0.5
mmol/ L vitamin B12, 7 mmol/ L tianium(III) citrate, 234
mmol/ L methanol, and 0.153 mmol/ L 2,3,4,5,6-PeCB. Con-
trols lacked either vitamin B12 or titanium(III) citrate. Ampules
were incubated in the dark at 20 °C.
Sedim ent-Sorbed 2,3,4,5,6-PeCB Am pule Experim ent.
A 1 L sediment sample was collected from the bank of the
Willamette River in Corvallis, OR. The sediment was spiked
with 10 mg of 2,3,4,5,6-PeCB and allowed to equilibrate for
8 months at 5 °C. The sediment was then sieved through a
mesh screen (1.5 mm pore size) to a final composition of
approximately 42% sand, 41% silt, and 17% clay. The chemical
properties of the sediment were as follows: pH 8.2 (pore
water), 6.2% organic matter (loss on ignition), 300 mg/ kg
iron, 200 mg/ kg manganese, and 9.8 ( 4.0 mg of 2,3,4,5,6-
PeCB/ kg of sediment. Deionized water was added to the
sediment to produce a slurry that could be easily mixed with
a magnetic stirrer.
Over 100 replicate ampules were prepared with 1 mL of
the vitamin B12 stock solution. The PeCB-spiked sediment
was homogenized with a magnetic stirrer, and 25 µL of the
slurry was added to each ampule with a repeating pipet.
After purging with nitrogen, 500 µL of titanium(III) citrate
was added, and the ampule was sealed. Each ampule
contained 0.33 mmol/ L vitamin B12, 80 mmol/ L titanium(III)
citrate, and 0.04 µg of 2,3,4,5,6-PeCB. Controls lacking vitamin
Results and Discussion
Vitam in B12s-Mediated Reductive Dechlorination of Aque-
ous and Sedim ent-Sorbed 2,3,4,5,6-PeCB. The reductive
dechlorination of aqueous and sediment-sorbed 2,3,4,5,6-
pentachlorobiphenyl (2,3,4,5,6-PeCB) by titanium citrate-
reduced vitamin B12s was evaluated over several months in
experiments conducted in hermetically sealed glass ampules.
Because ampules were created individually, PCB concentra-
tions were converted to molar fractions to reduce variability
between samples.
Aqueous 2,3,4,5,6-PeCB Am pule Experim ent. For the
aqueous ampule experiment, 40 replicate test ampules were
created, each containing 50 µg of 2,3,4,5,6-PeCB, 0.5 mmol/ L
vitamin B12, and 7 mmol/ L titanium(III) citrate in 660 mmol/ L
Tris (pH 8.2). Two sets of controls were created, lacking either
the reductant (titanium citrate) or vitamin B12. Ampules were
incubated in the dark at 20 °C.
Ampules were sampled to record the progress curve for
2,3,4,5,6-PeCB reductive dechlorination (Figure 1); each
homologue or congener was reported as a molar fraction of
the total chlorobiphenyls measured in each sample. The
B12 and/ or titanium citrate were also prepared. The ampules
9
VOL. 33, NO. 6, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 8 5 9