Table 9. Concentrations (g/L) of various analytes found by HPLC for
standards stored at room temperature in vials in the dark.
Days*
NB
m-NT
DNB
2,6-DNT 2,4-DNT
TNB
TNT
RDX HMX
Found concentration using HPLC
Replicate 1
0
121
554
50.2
54.5
52.7
57.4
52.1
101
489
1
122
566
50.1
54.2
52.2
55.5
37.9
96.7
482
4
123
550
51.9
53.1
53.3
38.9
24.3
106
458
Replicate 2
0
120
555
50.9
52.7
51.8
57.8
50.5
97.6
479
1
120
548
48.5
50.4
50.6
58.4
44.2
94.0
468
4
121
539
50.8
50.5
52.3
44.7
35.7
103
568
Target concentration
119
541
50.6
50.8
51.2
50.4
50.1
100
501
* Time at room temperature.
Figure 5. Effect of storage temperature on 10 g/L TNT calibra-
tion standard.
g/L) in acetonitrile containing water at a con-
vents, this water is removed from the solvent
centration of 10%. We compared the GC response
with anhydrous sodium sulfate. As stated above,
of standards with and without water by linear
we wanted to develop a method in which the
regression (Fig. 6 shows data for TNT). Slopes
sample extract could be analyzed by both GC and
ranged from 0.977 to 1.145, bracketing the ex-
by RP-HPLC using Method 8330. Acetonitrile is
pected value of 1.00. However, repeated injec-
the most efficient solvent for extraction of the
tions of standards containing 10% water resulted
Method 8330 analytes, especially the nitramines.
in tailing peaks and decreased peak heights of the
Polar solvents such as acetonitrile are not readily
nitramines and the amino compounds. We sus-
dried.
We prepared acetonitrile solutions with water
layer of the injection-port liner. Therefore air-
concentrations of 0 to 20% and analyte concentra-
tions of 25 g/L DNB, 2,6-DNT, 2,4-DNT, TNB,
drying of the solid phase prior to elution with
TNT, 4-Am-DNT, 2-Am-DNT, 50 g/L RDX, and
acetonitrile is important if the extract is to be
250 g/L HMX (Table 10). Using blocked ANOVA
analyzed by GC.
to compare mean peak heights, we found no sig-
nificant difference between 0%, 10%, 15%, and
Initial spike recovery
20% water. A small difference was observed for
5% water.
from water for many years (Jenkins et al. 1992).
We also made a series of standards (2.5 to 30
11