for the analysis of explosives. We tested numerous
PETN, which has a vapor pressure almost identical
chromatographic conditions and found that the
column was suitable for the analysis of NB, the ni-
contaminated sites, RDX is by far the more com-
trotoluenes, DNB, and the DNTs. Resolution of the
monly found of these two analytes. Further work
with mid-range polarity columns, described
other analytes was poor, the peaks for TNB, TNT,
below, resolved PETN from RDX, but led to co-
and RDX were uncharacteristically small, and
elutions with other analytes.
HMX did not elute at all. Here again, large column
We experimented with different column temper-
internal surface area, not total time in the GC, may
ature programs and injected a high-concentration
contribute to HMX loss.
solution of HMX onto the 15-m column. With a
high-temperature (250C) isothermal run, HMX
Confirmation columns
eluted as a broad jagged peak on the 15-m column.
We tested four 0.53-mm-i.d. columns for suit-
We next shortened the GC column to 6 m, as sug-
ability as confirmation columns. In order of
gested by Hable et al. (1991) and found that HMX
increasing polarity, these columns were J and W
now eluted as a sharp peak (Fig. 1). This dramatic
DB-1301 (6% cyanopropylphenyl methylpolysilox-
improvement was not due to total time in the GC;
ane), J and W DB17 (50% phenyl methylpoly-
rather, the decreased column length exposed the
siloxane), Restek RTX-200 (Crossbond trifluoro-
propyl methylpolysiloxane), and Restek RTX-225
analyte to less surface area. Tamiri and Zitrin
(50% cyanopropylmethyl50% phenyl methyl-
(1986) reported similar results when they ob-
served that PETN and RDX failed to elute intact
polysiloxane). The DB1301 was not acceptable be-
from a 30-m column but did elute from a 15-m col-
cause TNB coeluted with TNT, and DNB co-eluted
umn. Thus length of a typical GC capillary column
with 2,6-DNT. The DB17 was not suitable because
is an important consideration for successful chro-
TNB coeluted with TNT. The Restek RTX-200 re-
matographic analysis of the most thermally labile
solved the 8330 analytes at low linear velocity, but
explosives.
HMX is not detected (Fig. 2a). At high linear veloc-
We also tested an Alltech (Deerfield, Illinois)
ity, HMX was detected (Fig. 2b, Table 5), but PETN
MultiCapillary SE54 (5% phenyl methylpolysilox-
coeluted with RDX and 2-Am-DNT, and DNA was
ane) column. These columns are only 1 m long and
not resolved from 4-Am-DNT. Finally, on the RTX-
are composed of a bundle of over 900 liquid-phase
225, tetryl coeluted with RDX, and HMX was not
coated 40-m capillaries. They provide rapid analy-
detected (Fig. 2c). However, in subsequent analysis
sis of pesticides, and accommodate high carrier gas
of well-water samples from Louisiana AAP, we
velocities, so we reasoned they might be suitable
found this column to be excellent for confirmation
Table 5. Retention times (minutes) obtained for analytical and con-
firmation columns using temperature programs given in Figures 1
and 2.
DB-1
DB-1
DB-1
RTX-200
RTX-200
RTX-225
LV = 126 LV = 76
LV = 44
LV = 40
LV = 122
LV = 108
Analyte
(cm/s)
(cm/s)
(cm/s)
(cm/s)
(cm/s)
(cm/s)
NB
0.32
0.63
1.38
2.15
o-NT
0.47
0.92
2.06
2.78
0.95
m-NT
0.57
1.12
2.47
3.40
1.20
p-NT
0.62
1.22
2.69
3.72
1.40
NG
1.18
2.00
3.84
8.57
0.52
6.25
DNB
1.84
3.18
5.05
9.01
0.63
5.86
2,6-DNT
2.07
3.42
5.28
8.51
0.55
5.50
2,4-DNT
2.88
4.22
6.12
10.64
0.88
6.51
TNB
4.19
5.50
7.42
18.90
1.98
9.99
TNT
4.61
5.91
7.82
17.81
1.86
9.51
PETN
5.62
6.83
8.79
28.52
2.74
11.57
RDX
5.62
6.89
8.83
29.19
2.86
13.66
4-Am-2,6-DNT
6.77
8.02
9.92
23.80
2.45
12.65
3,5-DNA
6.83
8.11
10.07
26.08
2.65
13.32
2-Am-4,6-DNT
7.17
8.45
10.38
28.57
2.85
13.17
Tetryl
8.05
9.34
11.26
32.11
3.54
13.65
HMX
11.21
12.50
13.92
not eluted
6.29
not eluted
6
Previous Page
