Table 1. Retention times for various separations used
Table 2. Retention times for vari-
at CRREL.
ous separations used at WES.
Retention time (min)
Retention time (min)
Analyte
LC-18*
LC-CN*
Analyte
LC-18*
LC-CN*
HMX
0.3
13.0
HMX
2.4
1.4
RDX
0.1
8.3
RDX
3.5
7.3
TNB
0.7
5.2
TNB
4.6
4.3
DNB
0.3
5.2
DNB
5.6
4.4
3,5-DNA
9.2
6.6
3,5-DNA
6.1
5.6
NB
9.8
4.6
tetryl
6.2
9.2
TNT
1.2
6.3
NB
6.6
4.0
4ADNT
2.4
6.9
TNT
7.4
5.2
2ADNT
2.4
7.4
4ADNT
8.0
6.0
2,6-DNT
3.3
5.8
2ADNT
8.4
6.4
2,4-DNT
3.6
6.1
2,6-DNT
8.8
4.9
2,4-DNT
8.9
5.2
*Separations were conducted at 1.2
2NT
0.6
4.5
mL/min with an eluent of 1:1 metha-
4NT
1.8
4.7
nol/water.
3NT
2.4
4.8
tetryl breakdown product
4.6
8.1
2-amino-4-nitrotoluene
5.6
4.2
3-nitroaniline
4.2
3.8
2,4-diamino-6-nitrotoluene
4.0
6.3
2,6-diamino-4-nitrotoluene
2.1
4.8
4-amino-2-nitrotoluene
8.1
4.3
*Separations were conducted at 1.5 mL/min with an eluent
of 1:1 methanol/water.
1992) or were preconcentrated using either salt-
EXPERIMENTAL DESIGN
ing-out solvent extraction (SOE), cartridge solid-
AND RESULTS
phase extraction (SPE-C), or membrane solid-
phase extraction (SPE-M) (Jenkins et al. 1992).
Selection of initial test matrix
Primary analysis was conducted on a 25-cm
In a previous study (Grant et al. 1993), the
4.6-mm (5-m) LC-18 column (Supelco) eluted
preextraction holding times for nitroaromatics and
with 1:1 methanol/water (v/v) at 1.5 mL/min at
nitramines in water were evaluated using three
CRREL or 1.2 mL/min at WES (EPA 1992). At
sample matrices: reagent-grade water, groundwa-
CRREL, samples were introduced by overfilling a
ter from Enfield, New Hampshire, and surface
100-L sampling loop. At WES, 50-L samples
water from the Connecticut River in West Leba-
were introduced using a 200-L loop. Retention
non, New Hampshire. Of these, the most rapid
times of the analytes of interest are shown in Tables
rate of analyte transformation occurred in the Con-
1 and 2 for analyses conducted at CRREL and
necticut River water. After only seven days of re-
WES, respectively. Concentration estimates were
frigerator storage of Connecticut River water for-
obtained for most analytes from peak heights from
tified with TNB and TNT, only 45% of the initially
the digital integrator. In some instances, particu-
fortified TNB and 70% of the initially fortified TNT
larly for samples preconcentrated using Porapak
remained. The expected microbiological transfor-
RDX SPE cartridges, better quantitative results
mation products, 3,5-DNA from TNB and 2ADNT
were obtained using peak areas. At CRREL, the
and 4ADNT from TNT (Won et al. 1974, McCor-
identities of transformation products were con-
mick et al. 1976), appeared as the concentrations
firmed by analysis of some of the samples on LC-
of TNB and TNT decreased with time. Thus, forti-
CN, using a 25-cm 4.6-mm (5-m) LC-CN col-
fied Connecticut River water appears to be a good
umn from Supelco eluted with 1:1 methanol/wa-
choice as a test matrix for evaluation of alternative
ter (v/v) at 1.5 mL/min (EPA 1992). At WES, con-
preservation techniques.
firmation of analyte identities as well as quantita-
To ensure that Connecticut River water in the
tive results for 2ADNT and 4ADNT were obtained
vicinity of West Lebanon, New Hampshire, could
on an LC-CN column (Supelco) eluted with 1:1
be relied upon to be degradative for nitroaromatics,
methanol/water at 1.2 mL/min.
a short holding-time study was conducted in a
4
Previous Page
