levels of HMX. However, this cross-reactivity was
Short-range heterogeneity
not sufficient to allow it to be exploited to esti-
In the next phase of the study, soil samples
mate the concentration of HMX, since even high
were collected from subgrid D2B to compare the
concentrations of HMX such as 593 mg/kg (D2
short-range heterogeneity of contaminant distri-
location 1, surface) does not produce a D TECH
bution to that documented in our earlier study of
response higher than 4.05.0 mg/kg with an in-
TNT, DNT and ammonium picrate (Jenkins et al.
trinsic concentration of RDX of 1.56 mg/kg for
1996). Using a sampling protocol identical to the
the same sample. Roughly, the ratio of cross reac-
one used in the earlier study, we collected seven
2 order of mag-
tivity can be estimated as being 10
discrete samples were collected as shown in Fig-
nitude (Table 1). Using the RDX D TECH EIA test
ure 5 for subgrid D2B. These discrete samples
kit for the determination of HMX would lead to a
were homogenized and analyzed in duplicate,
sensitivity decrease of about a factor of 100, mean-
and a composite was prepared and analyzed in
ing a detection limit of approximately 50 mg/kg,
triplicate. The results presented in Table 2 and
which was not acceptable for our purposes.
Figure 5 include those from analysis using the on-
Results from the analysis of these samples for
site colorimetric method, those from laboratory
HMX and TNT are also presented in Table 1. In
HPLC analysis with the same acetone extracts
used for on-site analysis, and those from separate
all cases, HMX was present at concentrations about
subsamples analyzed at a commercial laboratory
two orders of magnitude greater than TNT, even
though the munition fired at this site
had a 70:30 ratio of HMX to TNT. The Table 2. Results from short-range spatial heterogeneity study
higher retention of HMX in these near for HMX with soil samples from subgrid D2B.
surface soils is probably due to its much
lower water solubility (5 mg/L vs. 150
HMX (mg/kg)
mg/L at 25C, Burrows et al. 1989) and Sample no.
Replicate
On-site
Acetone HPLC Method 8330
slower rate of dissolution. In addition,
Discrete samples
HMX is much less subject to biotrans-
1
a
101
112
220
formation or aerobic degradation than
b
99
109
120
is TNT (Grant et al. 1995), and indeed,
mean
100
111
170
concentrations of the monoamino trans-
2
a
20
18
17
formation products of TNT were often
b
11
13
6.7
present in concentrations equivalent to
mean
16
16
12
those of TNT (App. A).
3
a
198
191
320
HMX was present in samples from
b
168
189
250
grid D2, next to tank D, at concentra-
mean
183
190
285
tions about 10 times greater than for
4
a
115
144
150
those from grid D5, which was about 9
b
107
139
120
m from the target. With respect to depth,
mean
111
142
135
HMX concentrations were highest in the
5
a
313
318
290
surface samples (over 500 mg/kg for
b
329
337
310
grid D2 and about 100 mg/kg for D5)
mean
321
327
300
followed by the 07.5-cm segment. HMX
6
a
329
332
390
concentrations in soil samples from the
b
319
317
420
7.515-cm depth were generally a factor
mean
324
324
405
of 5 or more lower than in the 07.5-cm
7
a
46
69
110
depth soils, ranging from about 50 mg/
b
62
82
54
kg for samples from D2 to about 5 mg/
mean
54
75
82
kg for samples from D5. These results
confirmed what was observed in an ear-
Discrete sample mean
158
169
198
lier study where the depth of contami- Composite samples
nation was investigated (Thiboutot et
a
161
182
240
al. in press). Thus for this firing range,
b
182
180
120
c
185
190
locating HMX contamination sources
can be apparently accomplished best
Composite sample mean
176
184
180
using near-surface soil samples.
11
Previous Page
