Table 5. Results of direct analysis of preserved (pH 2) and
held prior to analysis. Since various samples
unpreserved Connecticut River water (six replicates each) us-
are analyzed at different times after filtra-
ing the direct analysis procedure of SW846 Method 8330.
tion, the size of this interference varies from
analysis to analysis. We did not fully appre-
Preserved
ciate or understand the significance of this
X
Analyte
(pH 2)
S
F*
t*
problem until after nearly all of the experi-
HMX
No
64.6
0.33
2.55
0.82
mental work in this report was completed,
Yes
66.2
0.83
including the evaluation using explosives-
RDX
No
00.5
0.94
1.89
1.20
contaminated groundwater samples. This
Yes
02.8
1.30
interference can be eliminated by the use of
TNB
No
52.7
0.63
2.77
0.55
Yes
53.4
0.97
disposable plastic syringes that do not have
DNB
No
39.9
0.52
4.98
0.03
a rubber-tipped plunger. This interference
Yes
39.9
0.23
will be discussed again in a later section on
3,5-DNA
No
59.5
0.32
4.49
3.04
the results of acidification on field-contami-
Yes
58.2
0.15
tetryl
No
41.2
0.58
2.28
0.61
nated groundwaters from the Naval Sur-
Yes
40.4
0.87
face Warfare Center.
NB
No
43.6
0.48
1.74
0.40
Since variances, except for TNT and
Yes
43.2
0.63
2ADNT, were not significantly different, fur-
TNT
No
91.4
0.37
31.8
1.61†
Yes
92.9
2.05
ther analysis of these results was conducted
4ADNT
No
57.2
0.42
4.14
0.21
by pooling the variances obtained for the
Yes
57.4
0.86
acidified and unacidified samples for a given
2ADNT
No
79.5
0.02
5.68
3.43††
analyte and performing a t test to assess
Yes
77.8
0.43
2,6-DNT
No
99.2
0.34
2.40
0.65
whether the mean values obtained for the
Yes
97.2
2.08
acidified and unacidified solutions were sta-
2,4-DNT
No
69.7
0.38
2.39
0.29
tistically different. Since the variances were
Yes
69.9
0.59
significantly different for TNT and 2ADNT,
2NT
No
74.4
0.69
2.12
0.21
Yes
74.1
1.01
variances were not pooled and t values were
4NT
No
78.4
0.97
3.17
0.80
calculated using the approach presented in
Yes
76.5
1.73
Miller and Miller (1984). Except for 2ADNT
3NT
No
59.9
0.95
3.37
0.06
and 3,5-DNA, means of acidified and
Yes
60.1
1.06
unacidified samples were found to be "not
X = mean; S = standard deviation
significantly different" at the 95% confidence
* Critical values for F and t at the 95% confidence level are 5.05 and
level. Thus, for these 13 analytes, there is no
2.23, respectively.
indication of a measurable effect on the ac-
† Critical value for t is 2.78 (d.f. = 5), thus the t value for TNT is not
curacy of the direct analysis method under
significant at the 95% confidence level.
†† Critical value for t is 2.37 (d.f. = 7), thus the t value for 2ADNT is
Method 8330 due to acidification of samples
significant at the 95% confidence level.
to pH 2. The significant differences in means
for the 2ADNT and 3,5-DNA were not sur-
found due to acidification. For 2ADNT, the F ratio
prising in light of the problems discussed above
was only 5.68, compared with a critical value of
for the amino compounds and others to be pre-
5.05, and this marginally significant value may be
sented later. Nevertheless, these differences are
a result of the unusually good precision obtained
too small to be of practical significance.
for this experiment. For TNT, the cause of the sig-
nificant F ratio was traced to a small and variable
Holding-time study using
coeluting peak in the chromatogram obtained for
pH 2 stabilization for
the acidified Connecticut River water. This peak
fortified river water
was not found in the unacidified water, nor was it
Based on the results above, acidification to pH
found in the methanol or the sodium bisulfate.
2 with sodium bisulfate was selected as the option
Subsequently this interference was found to be
with the greatest potential for conveniently stabi-
due to the use of a disposable syringe containing a
lizing waters containing nitroaromatics and
rubber-tipped plunger for sample filtration prior
nitramines. To evaluate this procedure more thor-
to RP-HPLC analysis. This interference is only pro-
oughly, a replicated holding-time study was con-
duced with acidified samples, and the degree of
ducted using all 15 target analytes (14 current
interference increases with time as samples are
SW846 Method 8330 target analytes plus 3,5-DNA).
11
Previous Page
