Table 2. Comparison of picric acid determinations
80
Soil Method (g/g)
between field screening method and HPLC.
Water Method (g/L)
Picric acid
Picric acid
Tetryl
60
screening
HPLC
HPLC
(g/g)
(g/g)
(g/g)
Sample
Crane162
0.0
0.0
0.0
Crane541
0.0
0.0
0.0
40
Hawthorne13
25,200
23,000
0.0
Mead42
1.9
1.0
60
Mead43
14.0
24.4
1265
20
Mead45
6.8
9.8
397
and Burlinson 1988, Jenkins et al. 1995), it is to be
expected in environmental samples contaminated
0
20
40
60
80
with tetryl.
Picric Acid in Soil and Water (RPHPLC)
Figure 9. Comparison between field methods and
Water method
RP-HPLC determination of picrate.
The MDL determinations were performed us-
ing fortified well waters, which produced a slightly
brown acidic extract. As a worst-case example of
background interference due to dissolved humic
color upon addition of water indicated that little
substances, this method was tested on river water
or no picrate was present. Subsequent RP-HPLC
that had been fortified with 125 g/L of tetryl and
analysis confirmed the absence of picrate in all of
allowed to sit at room temperature in the dark for
the samples.
60 days. HPLC analysis showed that 64% of the
A comparison between field method and RP-
tetryl had degraded, producing 25 g/L of picrate
HPLC determinations of picrate in contaminated
plus several other unidentified products. The An-
or fortified environmental matrices shows reason-
ion membrane turned a dirty brown color when
able agreement (Fig. 9). Rigorous statistical evalu-
this sample was extracted. Application of a drop
ations can be made after more field samples are
of EnSys reagent did produce a rusty colored spot.
analyzed. Future tests of the water method will
Elution with 5 mL of 10% H2SO4methanol pro-
have to be conducted as part of a sampling plan
duced a dark yellow-brown extract with an initial
that considers the 2-L sample requirement.
absorbance at 400 nm of 1.3. The subsequent dilu-
tion with reagent-grade water and the final absor-
Coordination with
bance reading resulted in a calculated concentra-
existing methods
tion of 39 g/L of picric acid.
The soil method can be easily added to accepted
Water samples from 34 monitoring wells at the
screening methods for the military explosives TNT
Naval Surface Warfare Center in Crane, Indiana,
and RDX in soil (Jenkins and Walsh 1992). A single
were screened for picrate using the proposed
100-mL acetone extract can be split for each of the
method. Since there was only one 500-mL sample
three tests. The Alumina-A cartridges are required
available from each well, samples were composited
for the RDX test to remove interfering nitrates and
to make 2-L samples for extraction. All compos-
nitrites.
ites had been previously extracted for nonionic
A field screening method for TNT and RDX in
explosives using Empore SDB extraction mem-
water that uses Empore SDB membranes has been
branes, which do not retain picrate. The samples
proposed (Jenkins et al. 1994). Both that method
at that point were colorless and free of sediment.
and the one proposed in this report require a
All the composites produced pink- or salmon-col-
vacuum filtration apparatus. A hand-operated
ored deposits on the Anion membranes. The addi-
vacuum pump can be used, but it requires con-
tion of EnSys reagent did not produce a darker
stant pumping for at least 20 minutes for sedi-
spot. These deposits were removed by the acidi-
ment-free samples. If sediment causes partial plug-
fied methanol, producing light yellow extracts
ging of the membrane, the hand pump is inad-
whose colors were reduced upon dilution with
equate. Realistically, a powered vacuum pump is
water. A spectrophotometer was not used for quan-
required to supply sufficient suction to extract 2-L
tification; however, the lack of a visible increase in
samples in a reasonable time.
9
Previous Page
