ples were collected from the crater and from 14 surface
ously extracted for 22 hours at a cycle time of six cycles
areas from immediately to the north of the crater to a
per hour. After cooling, the volume of solvent was meas-
distance of 76 ft (23 m) east of the crater, 50 ft (15 m)
ured and an aliquot removed for analysis. This sample
west of the crater, and 20 ft (6 m) south of the crater.
was labeled as SOX1.
Two samples (S-1E and S-1W) were collected just be-
Analysis of residues on steel plate
yond the area where visual deposition had occurred.
Once sampling was complete, we measured the
The surface of the steel plate was sampled using
extent of the surface that was visually contaminated
swiping techniques similar to that developed to esti-
with soot from the detonation. The impacted area is
mate the surface concentration of explosives residues
delineated by the shaded area in Figure 1. The total
on land mines (Leggett et al. 2000). Paper filter disks
(1.8 cm2) were soaked in methanol and placed randomly
surface area that was visually impacted was estimated
at 4100 ft2 (381 m2). The total area sampled within the
on the surface of the steel plate. Upon drying, the paper
visually impacted zone was 300 ft2 (28 m2), which was
disks were picked up with forceps and placed in an
7.3% of the impacted area.
amber vial containing acetone. Two samples of the plate
were analyzed. One sample utilized a single filter disk
Snow-sample processing
and 0.5 mL of acetone in the amber vial. A second sam-
The snow in the plastic bags was melted by placing
ple used five filter disks and 2.0 mL of acetone in the
the bags in the laboratory overnight. Small quantities
vial. The acetone extracts of the disks were initially
of ice remained in the bags in the morning, indicating
analyzed by gas chromatography-thermionic ionization
that the samples had remained at 0C during this peri-
detection (GC-TID) and quantitative results were
od. The water was black in color due to the presence of
obtained by gas chromatography-electron capture detec-
suspended soot particles. Once the last of the snow and
tion (GC-ECD) as described below.
ice melted, each of the samples was filtered to remove
GC-ECD determination
the soot by passing it through glass fiber filters (What-
man glass microfiber, 47 mm, grade GF/A). Depend-
All SOE and SOX samples were analyzed by GC-
ing on the amount of soot in the sample, it took as many
ECD on an HP 6890 gas chromatograph equipped with
a microcell Ni63 ECD (300C). We used direct injec-
as 14 individual filters for a given sample. These filters
tion (250C) of 1-L extracts in a packed port that was
were retained and were extracted separately, as de-
scribed below.
equipped with a deactivated Restek Uniliner. Primary
analysis was conducted on a 6-m- 0.32-mm-i.d. fused-
After filtration, each volume of snowmelt was meas-
silica, 1.5-m film thickness of 5%-(phenyl)-95%-
ured and a 1520-mL aliquot was placed in a 2-L volu-
metric flask containing 496 g of sodium chloride (Miy-
dimethyl polysiloxane RTX-5 column from Restek. The
ares and Jenkins 1991, EPA 1994). A magnetic stir bar
GC oven was temperature programmed as follows:
100C for 2 min, 10C/min ramp to 250C, 3-min hold.
was added and the flask stirred to dissolve the salt. A
325-mL aliquot of acetonitrile (ACN) was added and
The carrier gas was helium at 10 mL/min (linear veloc-
the flask was stirred vigorously for 30 minutes. The
ity about 100 cm/sec). The makeup gas was nitrogen
magnetic stirrer was then turned off and the phases were
(40 mL/min). Further details of the procedure, including
allowed to separate for 30 minutes. The ACN phase on
conditions for the confirmation column, may be found in
the top, about 25 mL, was removed and the volume
SW-846 Method 8095 (Walsh and Ranney 1998, EPA
measured using a graduated cylinder. This first extract
1999). Retention times for the target signatures are pre-
was labeled salting-out extract 1 (SOE1). An additional
sented in Table 1. Detection limits reported by Walsh
20 mL of acetonitrile was then added to the volumetric
and Ranney (1998) for a method that utilized solid phase
extraction rather than salting-out extraction are shown
flask and the flask was again stirred vigorously for 30
in Table 2. All analyte concentrations were within the
minutes. The magnetic stirrer was turned off again and
linear range of the ECD and concentrations reported
the phases allowed to separate. The second ACN extract
were taken from the determination on the primary col-
was then removed, the volume measured (about 20 mL),
umn (RTX-5).
and it was labeled SOE2.
To ensure that we included any explosives residues
GC-TID for acetone extracts of steel plate
that were deposited on the inside of the bags, the sam-
The two acetone extracts of the surface of the steel
ple bags were wiped with a Whatman filter and the filter
plate were analyzed using GC-TID according to meth-
was placed in a Soxhlet extraction thimble. The filters
ods developed elsewhere (Hewitt and Jenkins 2000).
used to remove soot from that sample were added to
TNT, RDX, and 2,4-DNT were detected. Analytes were
the thimble, which was placed inside a Soxhlet extrac-
then quantified using the GC-ECD method described
tor. A 250-mL aliquot of ACN was added to the receiver
above.
of the Soxhlet extractor and the samples were continu-
3
Previous Page
