vironmental Center. Field-contaminated soils from
OH
CH3
O2N
CH 2 CH 2 CH 3
Hawthorne, Nevada; Naval Surface Warfare Cen-
ter (NSWC), Crane, Indiana; and Nebraska Ordi-
nance Plant, Mead, Nebraska, were used to test
NO 2
the methods. Well waters from NSWC were also
Figure 2. Chemical structure of
used.
Dinoseb.
All solvents used for extraction, collection and
elution were HPLC-grade from Baker. Reagent-
normal-phase extraction material. When Dinoseb
grade water was prepared using a Milli-Q Type 1
is present, the white Florisil turns yellow. A visual
Reagent-Grade Water System (Millipore). The
detection limit of 5 g/g was reported; however,
solid-phase extraction (SPE) materials were Florisil
(Supelco), Alumina-A SPE cartridges (3 mL-1 g,
was not suggested.
Supelco) and Anion-Empore SPE membranes (47
mm, 3M-Varian). The syringe filters were Millex
SR (0.45 m, 25 mm, Millipore). Immunoassay kits
Objectives
for TNT were EnviroGard (Millipore) and D-TECH
The major goal of this effort was to develop a
field screening method for ammonium picrate and
(EM Science). The color reagent for nitroaromatics
picric acid that could be used in conjunction with
was from the TNT kit from EnSys (Research Tri-
field screening methods already established for
angle Park, North Carolina).
performed by RP-HPLC on a 25- 4.6-cm (5 m)
(Jenkins and Walsh 1992, Jenkins et al. 1994). If
detection limits and action levels are similar, quali-
LC-18 (Supelco) column. The analytes were eluted
tative identification and confirmation of picrate in
u s i n g 1.5 mL/min of 60:40 (v/v) aqueous
soil and water is a reasonable stopping point for a
buffer:methanol. The buffer was 0.05 M KH2PO4,
screening method. However, if the screening can
adjusted to pH 3.5 with acetic acid. The analytes
be semiquantitative or quantitative, then site man-
were detected at 365 nm for picrate and 254 nm
agers can be alerted to contaminant levels that
for all others. The spectrophotometer used for the
might produce a significant release if the site is
field method was a battery-powered Hach DR-
further disturbed. This is particularly true in the
2000 (Loveland, Colorado).
case of picrate, where it could conceivably be ex-
posed to water after having been sequestered in
clay or sheltered from rain by pavement or build-
RESULTS
ings. Furthermore, it is hoped that a rapid and
inexpensive field test will allow screening for
Field screening in soil
picrate at nonmilitary sites.
The strategy that was employed in this research
Extraction conditions
was to investigate the adaptability of some exist-
An initial experiment investigated the poten-
tial for adapting the Dinoseb/Florisil method
ing screening methods that detect related com-
(Anderson et al. 1993) for the detection of ammo-
pounds and to develop new methods based on
nium picrate and picric acid. When picric acid was
the behavior of picric acid as it changes from a
dissolved in MeCl2, it produced a colorless solu-
colorless, undissociated acid to a yellow picrate
tion characteristic of solvated, but undissociated,
ion. The other characteristic of picrate that might
picric acid (Majersky and Dybalova 1975). When
allow selective separation from other yellow com-
white Florisil powder was added, the solution re-
pounds discussed below is its high acidity. Ion
mained colorless while the powder turned a bril-
exchange materials and binding and elution con-
liant yellow. The color was due to the formation of
ditions were sought that would retain, then release,
picrate ion pairs on the basic sorption sites of the
Florisil.
The application of this ion-exchange reaction
to soil samples was then investigated using 2 g of
EXPERIMENTAL METHODS
soil mixed with 5 mL of MeCL2 in a 22-mL glass
vial. When wetted soils were extracted, phase sepa-
Analytes used for spikes were Standard Ana-
ration of the MeCL2 was difficult to achieve if there
lytical Reference Material from the U.S. Army En-
3