linear regression. The data show that the response
lected from the top of the columns. The cap of the
is linear up to 5800 g/L (Fig. 5).
column was removed and a scintillation vial was
lowered down to the center of the fluidized bed on
Daily calibration
the end of a heavy gauge copper wire. The vial was
A daily calibration standard was prepared by
raised slowly through the fluidized bed to allow the
injecting 20 L of the low stock standard into 10 mL
maximum amount of clay to be collected. The vial
of water as described above. The vial was then
was removed from the wire and immediately
shaken vigorously, by hand, for 1 minute to estab-
capped with an aluminum-faced screw cap. A sin-
lish the equilibrium headspace concentration. A
gle clay sample was collected at each sampling
sufficient number of replicates was prepared such
event.
that three standards were analyzed at the start of
the analysis, and additional replicates were anal-
Sample preparation
yzed after every sixth sample and as the final sam-
Water samples were prepared for headspace
ple of the analysis. A peak height response for the
analysis by removing a 12-mL aliquot of water from
TCE was recorded for each analysis and the mean
the vial with a needle and syringe. The needle was
peak height response calculated. The daily response
inserted into the vial to allow air to be drawn in
factor was calculated by taking the ratio of the
while the syringe removed water, thus creating a
mean peak height response over the concentration
headspace at atmospheric pressure. The vial was
of the standard. The resulting units for the response
then shaken vigorously, by hand, for 1 minute to
factor were peak height per microgram per liter.
establish the equilibrium headspace concentration.
Note that, during the remediation phase, the 12-mL
Sample collection
aliquot removed from the vial was transferred to a
We collected water samples in 22-mL headspace
scintillation vial and held for subsequent phenol
autosampler vials. The vials were overfilled with
analysis. After the headspace analysis was per-
the water sample, then capped with an open top
formed, the sample was held for subsequent chlo-
crimp cap and a gray butyl, Teflon-lined septum.
ride analysis.
During the loading process, influent and efflu-
To analyze the clay samples, a 1-g subsample
was removed from the collection vial using a 3-cm3
ent water samples were collected. The influent wa-
modified plastic syringe. The syringe was modified
ter was collected before it passed through the
by cutting off the Luer Lok tip and removing the
packed adsorption bed and the effluent water was
rubber tip from the plunger. In the syringe 1 g of the
collected after. The first hour, samples were col-
wet clay material was equivalent to approximately
lected at time 0, 15, 30, 45, and 60 minutes. For the
1.5 cm3. The subsample was transferred to a second
next 2 hours, samples were collected every 30 min-
scintillation vial that contained 1.0 mL of methanol.
utes, then every hour up to 7 hours. On subsequent
This vial was then capped and placed in an ultra-
days, samples were collected every 2 hours during
sonic bath for 10 minutes. The methanol was re-
the work day (four times).
moved and transferred to another vial. A second
Sample collection continued until the columns
1.0-mL aliquot of methanol was added to the clay
were determined to be loaded. The influent sam-
sample and the sample was again sonicated for 10
ples were collected directly from the holding tank
minutes. Two additional 10-minute extractions
by submerging the vials with the open end pointed
with 1.0 mL of methanol were performed (four ex-
upwards. The effluent samples were collected from
tractions in total). The methanol extracts were all
a sampling port at the top of the columns. When
combined in one vial. The vial was centrifuged for 2
sampling, we closed the drain port to raise the level
minutes at 2500 rpm to separate the suspended sol-
of the water in the column above the sampling port.
ids from the liquid. Three 100-L aliquots of the
This was done to ensure that samples were not
extract were each transferred via syringe to previ-
being collected at the water/air interface where the
ously sealed, 22-mL headspace autosampler vials,
concentration could be varying because of volatil-
each containing 10 mL of water. This vial was shak-
ization. Upon collection of the samples, the drain
en vigorously by hand for 1 minute to establish the
port was reopened. During the remediation phase,
equilibrium headspace concentration.
water samples were collected daily in a way identi-
cal to that used for the effluent samples. All water
Sample analysis
samples were collected in triplicate.
All headspace standards and samples were anal-
Clay samples were collected on day zero, then
yzed using the headspace GC system described
every third day, in the morning. Samples were col-
6
Previous Page
