1983, McCormick et al. 1984), but biodegradation
vided concentration estimates that were essen-
of both are thought to be primarily anaerobic pro-
tially equivalent to those from the standard labo-
cesses.
ratory method. Linear correlation analysis for
HMX results from the on-site colorimetric method
with a laboratory HPLC method using the same
Analysis of explosives residues
acetone extract resulted in a slope of 1.008 and a
remaining on metallic debris
Although the site had been "cleared" by safety
correlation coefficient of 0.992; the intercept was
personnel prior to site characterization, the ground
small and nonsignificant at the 95% confidence
was littered with small pieces of metallic debris
level. Correlations of the on-site HMX results with
from detonated projectiles, empty rocket bodies,
those from Method 8330 conducted at an inde-
fins and booster cups. A 7-g sample of metallic
pendent commercial laboratory were also good,
debris was collected and subsequently extracted
but less so. The slope of the model with intercept
with acetone in the CRREL laboratory. RP-HPLC
for this analysis was 1.016 with an intercept of
analysis revealed the presence of HMX and TNT
25.0 and a correlation coefficient of 0.972. The
at 50 mg/kg and 0.1 mg/kg, respectively. The
poorer relationship found here was attributed to
higher residual HMX is probably due to its lower
the analysis of separate subsamples of soil, rather
aqueous solubility and rate of dissolution com-
than the analysis of aliquots of the same extract
pared with TNT. The concentrations associated
used in the former comparison.
with this debris do not seem to be high enough to
For TNT, a similar correlation of the on-site
be the major mode of contamination for the soils
colorimetric result with that from HPLC analysis
at CFB-Valcartier, but does demonstrate that not
of the same acetone extract resulted in a slope of
all the explosive charge is consumed during deto-
1.05, and intercept of 0.63 (significant at the 95%
nation.
confidence level), and a correlation coefficient of
0.975. No correlation with Method 8330 results
was obtained due to substantial variability en-
GC comparisons
The GC proved to be useful for analyte confir-
countered among subsamples for TNT, which was
mation following HPLC analysis, especially for
present at much lower concentrations than HMX.
confirmation of DNT. For example, when acetone
This result indicates that care should be used when
extracts were analyzed by HPLC using the LC-
comparing results from on-site and laboratory
CN column, a peak eluting just prior to the reten-
analyses conducted with separate subsamples,
tion time for TNT and corresponding to the reten-
unless soils are homogenized carefully before
tion time for DNT was observed in several samples
splitting.
(Fig. 8). Because of the close retention times for
It is important to remember that the results from
these two analytes on LC-CN, the analytes are not
these on-site colorimetric methods were available
resolved if one analyte is at a much higher con-
to site investigators the day following soil sample
centration than the other. Based on site character-
collection, allowing their use to adjust site activi-
ization conducted prior to this study, the pres-
ties to reflect an increasing knowledge of con-
ence of DNT was not anticipated; hence confir-
taminant distribution. The use of EIA RDX tests
mation was required. Analysis of the acetone
was abandoned after the first set of samples, since
extracts by GC confirmed the presence of DNT,
HMX is present as the main contaminant and the
and permitted quantification of the lower concen-
cross-reactivity for this nitramine is not adequate
tration analyte (Fig. 9).
to provide sensitive detection. The comparison of
results from the D TECH EIA method and those
from acetone HPLC for TNT determination led to
a slope of the linear model of 0.69 with an inter-
SUMMARY AND CONCLUSIONS
cept of 1.24 and a correlation coefficient of 0.91,
which indicates a somewhat poorer correlation
On-site methods
The results of this study provide information
between the two methods compared to the excel-
in a number of areas that should be quite useful
lent general correlation observed between the colo-
for planning site characterization activities at ex-
rimetric and acetone HPLC methods.
plosives-contaminated areas. First, the utility of
on-site methods for obtaining concentration esti-
Short-range spatial heterogeneity
mates for TNT and HMX is clearly demonstrated.
Short-range spatial heterogeneity of HMX con-
Colorimetric methods for HMX and TNT pro-
centrations at the CFB-Valcartier antitank range
24
Previous Page
