1,200
Ottawa Sand
50 g/kg
1,000
800
600
400
200
0
0
5
10
15
Time (min)
f) Ottawa sand, 50 g/kg.
Figure 5 (cont'd).
Table 2. Method detection limits (g/kg) of
and accuracy of an analytical method can produce
confounding results because of the differing
nitroaromatic, nitramine, and nitrate esters in
analyte stability in newly spiked versus field-con-
spiked soils determined by GC-ECD.
taminated aged soils (Grant et al. 1995). This dif-
ference became apparent when we spiked 25-g soil
Ottawa sand
AEC soil
aliquots with the analytes of interest for mine de-
Method 1
Method 1
Method 2
tection (Table 1). The first soil we spiked was a silt
obtained locally. As stated above, we added 1 mL
1,3-DNB
0.93
0.73
1.2
of spike solution to 25 g of soil to yield a bulk
1,4-DNB
0.86
analyte concentration of 2 g/kg. However, this
1,2-DNB
0.64
2,6-DNT
0.81
0.69
small volume of spike solution wetted only a small
2,4-DNT
0.88
0.69
0.86
portion of the soil, unlike when we spiked 2-g soil
TNB
4.91
1.6
aliquots. When we extracted the silt with acetoni-
TNT
0.45
2.4*
11*
trile containing 25 g/L 3,4-DNT, recovery of all
RDX
3.1
3.4
2.6
analytes was nil. The peak height for 3,4-DNT, the
4-Am-DNT
1.6
1.5
0.70
3,5-DNA
2.0
2.1
internal standard present throughout the extrac-
2-Am-DNT
3.1
2.0
0.84
tion process, was normal, thus the loss of the
NB
14
17
spiked analytes must have occurred during the
o-NT
12
12
one hour of aging between the addition of the
m-NT
11
11
spike solution and the extraction solvent. To de-
p-NT
9.5
10
NG
31
13
termine whether the loss was matrix-specific, we
PETN
35
16
spiked duplicate 25-g samples of the following:
Tetryl
66†
20
glass beads (25-micron), Ottawa sand, AEC soil,
HMX
26
25
and the same silt where loss was observed initially.
Each matrix had interferences that co-eluted with
Method 1: 2 g soil extracted with 10 mL acetonitrile.
Method 2. 25 g soil extracted with 50 mL acetonitrile.
at least one analyte, but we found that recovery
*Ottawa sand contained an interference that co-eluted with
bracketed 100% for analytes spiked onto the glass
TNT.
beads and the Ottawa sand, bracketed 90% for one
†AEC soil contained an interference that co-eluted with
replicate of the AEC soil and 50% for the other rep-
tetryl.
17
Previous Page
