Evaluation of Clean Solid Phases for Extraction of
Nitroaromatics and Nitramines from Water
THOMAS F. JENKINS, PHILIP G. THORNE, KAREN F. MYERS,
ERIKA F. MCCORMICK, DON E. PARKER AND B. LYNN ESCALON
INTRODUCTION
traction (SPE) as an alternative to SOE and report-
ed excellent recovery and detection limits that were
One of the U.S. Army's most serious environ-
very similar to those for SOE. Winslow's results
mental problems is associated with sites contami-
were obtained using Porapak R, a divinylbenzene
nated with residues of secondary explosives. Con-
n-vinylpyrrolidone co-polymer, in the cartridge
tamination at these sites was chiefly caused by
format. LeBrun et al. (1993), using SPE in the mem-
manufacture of the explosives, loading of explosives
brane format, reported excellent recoveries of the
into ordnance, and disposal of off-specification or
analytes in Method 8330 using a membrane com-
out-of-date material. Residues from these activi-
posed of styrene-divinylbenzene. Recently Bouvier
ties contain the explosives, manufacturing impu-
and Oehrle (1995) reported on the use of Porapak
rities, and environmental transformation products
RDX for cartridge SPE preconcentration of nitroaro-
(Walsh et al. 1993). Unlike many other organic
matics and nitramines.
chemicals, these compounds are quite mobile in
Because of a number of potential advantages of
the soil and have resulted in serious groundwater
SPE over SOE, we conducted a three-way compari-
contamination (Kayser and Burlinson 1982, Pugh
son of SOE, cartridge-based SPE using Porapak R
1982, Rosenblatt 1986, Maskarinec et al. 1986,
(SPE-C), and membrane-based SPE (SPE-M) using
Spaulding and Fulton 1988). Plumes of contami-
styrene-divinylbenzene membranes (Empore SDVB)
nated groundwater, often miles in length, have
for preconcentration of waters containing nitroaro-
been identified at military sites, with some ex-
matics and nitramines (Jenkins et al. 1992, 1994).
tending beyond installation boundaries.
This evaluation included estimating detection
A number of laboratory methods have been
capability and analyte recovery using fortified
developed to characterize water samples poten-
reagent-grade water, and analyte recovery for a
tially contaminated with secondary explosives. At
series of field-contaminated groundwater samples
present, however, the method most often used by
from the U.S. Naval Surface Warfare Center
contract laboratories conducting analyses for the
(NSWC), Crane, Indiana. Overall, the results can
Army is SW846 Method 8330 (EPA 1992). This is
be summarized as follows:
a reversed-phase, high-performance liquid chro-
1) The three methods were comparable with
matographic (RP-HPLC) method that specifies 14
respect to low-concentration detection capability,
ranging from 0.05 to 0.30 g/L.
target nitroaromatic and nitramine analytes and
two protocols for water analysis. When detection
2) Percent recoveries generally exceeded 80%,
limits ranging between 4 and 14 g/L are ade-
except for HMX (octahydro1,3,5,7-tetranitro
quate for project requirements, a direct injection
1,3,5,7-tetrazocine) and RDX (hexahydro1,3,5-tri-
procedure can be used that does not require sam-
nitro1,3,5-triazine) by membrane SPE where recov-
ple preconcentration prior to RP-HPLC determin-
eries were lower.
ation. When lower detection limits are needed, a
3) Large interferences were found on about half
protocol including a salting-out solvent extraction
of the groundwater samples from the NSWC using
(SOE) preconcentration step is specified (Miyares
the two SPE methods but none were found by SOE.
and Jenkins 1990, 1991). Winslow et al. (1991,
4) The SPE interferences were traced to a matrix
1992) have proposed the use of solid-phase ex-
interaction of the SPE polymers with low pH
Previous Page
