was conducted on a 6-m 0.53-mm (or 0.32-mm) ID
of the ECD (about 200 g/L), extracts were analyzed
fused-silica, 1.5-m film thickness of 5%-(phenyl)-
by reversed phase high performance liquid chromatog-
95%-dimethyl polysiloxane (either HP-5 from Hewlett-
raphy with an ultra violet detector (RP-HPLC-UV) as
Packard, or a RTX-5 from Restek). The GC oven was
described below.
temperature programmed as follows: 100C for 2 min-
utes, 10C/min ramp to 250C, and 3-minute hold. The
RP-HPLC-UV determination
carrier gas was either hydrogen or helium at 12 to 15
Soil extracts with analyte concentrations above 200
g/L were often also analyzed by RP-HPLC-UV on a
mL/min (linear velocity about 100 cm/s). The makeup
gas was nitrogen (30 to 40 mL/min). If a peak was
modular system composed of a Spectra-Physics Model
observed in the retention window for a specific ERC
SP8800 ternary HPLC pump, a Spectra-Physics Spec-
compound, the extract was reanalyzed on one of the
tra 100 variable wavelength UV detector set at 254 nm
following two confirmation columns: Restek RTX-225
(cell path 1 cm), a Dynatech Model LC241 auto sam-
(50% cyanopropylmethyl, 50% phenyl methyl
pler equipped with a Rheodyne Model 7125 sample loop
polysiloxane) or Restek RTX-200 (crossbond
injector, and a Hewlett-Packard 3396A digital integra-
trifluoropropyl methyl polysiloxane). If the identity of
tor set to measure peak heights.
the peak was still in doubt after the first confirmation
Extracts were diluted with reagent grade water (0.50
column, the extract was reanalyzed on the second con-
mL acetonitrile extract and 2.00 mL of water). Separa-
tions were conducted on a 15-cm 3.9-mm (4 m)
firmation column. Further details of the procedure may
be found in SW-846 Method 8095 (U.S. Environmen-
NovaPak C-8 column (Waters) eluted with 85/15 wa-
tal Protection Agency 1998). Retention times for the
ter/isopropanol (v/v) at 1.4 mL/min. Samples were
target signatures are presented in Table 10. Detection
introduced by overfilling a 100-mL sampling loop.
limits reported by Walsh and Ranney (1999) are shown
Retention times of the analytes of interest are shown in
in Table 11.
Table 12. Concentrations were estimated against
If analyte concentrations were within the linear range
multianalyte standards prepared from standard analyti-
of the ECD, concentrations reported were taken from
cal reference materials (SARM). Quantitative results
the determination on the primary column, unless there
were obtained from peak heights on a variable wave-
appeared to be co-elution with another compound. In
length detector set at 254 nm.
such cases, reported concentrations were taken from
Sampling to estimate mine surface
determination on one of the confirmation columns. For
concentrations that appeared to be above the linear range
been buried for over 1 year
After the mines had been removed from the ground
to allow soil sampling under them in July and Novem-
Table 11. Method detection limits (g/kg) of
ber 1999, they were vigorously brushed to remove a
nitroaromatics and nitramines in soil deter-
majority of the adhered soil. However, it should be noted
mined by GC-ECD. (After Walsh and Ranney
that the mines still had some soil on their outside sur-
1999.)
faces, but we did not wash it off because of the possi-
Analyte
1,3-DNB
0.73
Table 12. RP-HPLC retention times (minutes)
1,4-DNB
0.86
for land mine signature compounds present
1,2-DNB
0.64
in high concentration.
2,6-DNT
0.69
2,4-DNT
0.69
Analyte
TNB
1.6
2,4,6-TNT
0.45
RDX
3.0
RDX
3.4
3-NA
3.5
4-Am-DNT
1.5
1,3-DNB
4.8
3,5-DNA
2.1
2,4,6-TNT
5.4
2-Am-DNT
2.0
2-Am-6-NT
5.9
NB
17
4-Am-2-NT
6.5
o-NT
12
2-Am-4-NT
6.9
m-NT
11
2,4-DNT
10.9
p-NT
10
2,6-DNT
13.4
Tetryl
20
4-Am-DNT
15.2
HMX
25
2-Am-DNT
17.2
3-NA
????
11
to contents