groundwaters that apparently caused the release
Teflon bailers. Samples were collected in 1-liter pre-
cleaned amber glass bottles and were stored and
of unreacted monomers or other contaminants
shipped at 4C.
from the interior of the polymeric materials.
At least partly in response to the problems
identified above, several manufacturers of SPE
RP-HPLC analysis
All water samples were analyzed by RP-HPLC
materials sought to improve the retention of SPE
at WES. Depending on the specific test conducted,
materials for very polar organics such as HMX
water samples were either analyzed using the
and RDX and experimented with new cleaning
direct method specified in SW846 Method 8330
procedures to better remove interferences from
(EPA 1992) or were preconcentrated using either
the SPE materials. As a result, Waters Corpora-
SOE, SPE-C, or SPE-M as described below (Jen-
tion released a new ultra-clean SPE material for
kins et al. 1992).
use in cartridge SPE under the name Porapak
Primary analysis was conducted on a 25-cm
RDX (Bouvier and Oehrle 1995), and 3M Corpora-
4.6-mm (5-m) LC18 column (Supelco) eluted
tion developed a new surface-modified styrene-
divinylbenzene membrane that also had been
with 1:1 methanol/water (v/v) at 1.2 mL/min.
Injection volume was 50 L introduced using a
cleaned more extensively (Empore SDB-RPS). Ini-
200-L sample loop. Concentration estimates were
tial tests at the U.S. Army Cold Regions Research
and Engineering Laboratory (CRREL) and else-
obtained from peak heights from a Waters 820
where indicated that these materials were indeed
Maxima chromatography workstation. The iden-
cleaner than the original SPE materials.
tity of target analytes and transformation products
was confirmed by analysis of the samples on a
25-cm 4.6-mm (5-m) LC-CN column from
OBJECTIVE
Supelco eluted with 1:1 methanol/water (v/v) at
The objective of this study was to reassess SPE
1.2 mL/min (EPA 1992). Quantitative results for
for preconcentration of nitroaromatic and nitramine
the 2-amino- and 4-aminodinitrotoluenes (2ADNT
explosives from water, using the newly released,
and 4ADNT) were also taken from the LC-CN de-
manufacturer-cleaned SPE materials. Special atten-
termination, since better separation of these two
tion was given to recovery of HMX and RDX, be-
analytes were obtained on this column. Retention
cause of the low recoveries found for these analytes
times of the analytes of interest for both separa-
with membrane SPE in the initial study. Assessment
tions are shown in Table 1.
of the level of contamination resulting from use of
Table 1. Retention times for various RP-
these manufacturer-cleaned materials was conducted
HPLC separations.
using both reagent water samples and some ground-
waters from the Naval Surface Warfare Center
Retention time (min)
Analyte
LC18* LC-CN*
low-pH waters that had revealed the contamination
HMX
2.4
11.4
problem with the initial SPE materials.
RDX
3.5
7.3
TNB
4.6
4.3
DNB
5.6
4.4
EXPERIMENTAL
3,5-DNA
6.1
5.6
tetryl
6.2
9.2
Conduct of study
NB
6.6
4.0
TNT
7.4
5.2
This work was jointly conducted by the U.S.
4ADNT
8.0
6.0
Army Engineer Waterways Experiment Station
2ADNT
8.4
6.4
(WES), Vicksburg, Mississippi, and CRREL, using
2,6-DNT
8.8
4.9
fortified reagent-grade water and actual ground-
2,4-DNT
8.9
5.2
2NT
10.6
4.5
water samples from the NSWC.
4NT
11.8
4.7
3NT
12.4
4.8
Collection of groundwater samples
tetryl breakdown product
4.6
8.1
Groundwater samples were taken with bailers
2-amino-4-nitrotoluene
5.6
4.2
3-nitroaniline
4.2
3.8
that were rinsed once with isopropyl alcohol and
2,4-diamino-6-nitrotoluene
4.0
6.3
three times with distilled water between samples.
2,6-diamino-4-nitrotoluene
2.1
4.8
Wells were purged with a PVC bailer to a depth
4-amino-2-nitrotoluene
8.1
4.3
midway down the well stream, allowed to re-
*Separations were conducted at 1.2 mL/min
charge a minimum of 2 hours, then sampled with
with an eluent of 1:1 methanol/water.
2
Previous Page
