OH
OH
O 2N
NO 2
O 2N
NO 2
O
O
+
CH2 C CH3
CH2 C CH 3
H
NO 2
NO 2
Colorless
Pink
Figure 8. Formation of the colored Meisenheimer anion from picric acid.
the required 50% dilution of the 5-mL eluent pro-
similar to the curve calculated for the soil method.
vided a background correction factor similar to
This became insignificant when absorbances were
the soil method.
rounded to the nearest 0.01. Concentrations of
picric acid should be reported using two signifi-
Qualitative confirmation of picrate
cant digits. For a 2-L water sample and a spectro-
The extraction of 2 L of reagent-grade water
photometer with a 1-cm path-length cell:
fortified with 2 g/L of picrate produced a visibly
picric acid (g/L) = 280 (g/L-ABS unit)
yellow surface on the Anion membrane. A lack of
(Final ABS @ 400 nm 0.5
any yellow color on the membrane is an indica-
tion that less than 2 g/L of picrate is present in
Initial ABS @ 400 nm) (Fig. 7).
the sample. If there is a yellow color remaining
after the methanol rinse, it could be picrate at a
To estimate the MDL, a series of seven replicate
level above 2 g/L. Since picric acid is a nitro-
2-L well-water samples fortified with 7.5 g/L of
aromatic, it forms a colored Meisenheimer anion
picric acid were extracted, rinsed with 5 mL of
when exposed to the Janowsky conditions of a
methanol, then eluted with 5 mL of 10% H2SO4
basic ketone solution (Kabeya et al. 1973) (Fig. 8).
methanol and the absorbance at 400 nm of the
Experiments were performed to see if this reac-
resulting extract recorded. The extract was diluted
tion would occur with the sorbed picrate. A qua-
with 5 mL of reagent-grade water and the absor-
ternary ammonium salt reagent commercially
bance at 400 nm of the diluted extract obtained.
The MDL was 3.6 g/L.
available as part of the EnSys TNT detection kit
was used to confirm the presence of picric acid on
A daily calibration standard is made by dilut-
ing 30 mL of a 10 g/mL aqueous solution of picric
the surface of the Anion membrane, both with and
without brown humic interferences. An additional
acid to 2 L with reagent-grade water and perform-
acetone rinse was added after the methanol rinse.
ing the method. The absorbance/cm at 400 nm
should be 0.56 0.03.
One or two drops of EnSys reagent was then ap-
plied and the color observed. When the Anion
membrane was yellow or brown, the Ensys re-
Performance evaluations
agent turned a noticeable pink or dark rust, re-
spectively, confirming that the yellow color was
Soil method
due to a nitroaromatic.
The method was tested on field-contaminated
This colorimetric confirmation scheme was tried
soils. The results are listed in Table 2. Soils from
for the soil method; however, in the cartridges the
Crane, Indiana, produced "straw-colored" acetone
yellow was dispersed throughout a few millime-
extracts. Analyses by HPLC showed that they con-
ters of the sorbant bed so that the pink that was
tained no picrate. The field screening method pro-
produced was barely discernible, even when there
duced a very light yellow Alumina-A extract that
were no brown interferences. The presence of any
was reduced by dilution. The soil from Hawthorne
brown completely obscured the pink EnSys color
required a 360-fold dilution to fall within the lin-
in the cartridge.
ear range of the calibration curve. The soils from
Mead had been analyzed previously by Method
Method detection limit
8330. The only detected analyte had been tetryl.
A standard curve was constructed by making a
Both the field screening method and the HPLC
series of 2-L solutions of reagent-grade water for-
method using the buffered eluent system revealed
tified with picric acid, plus a blank. The calculated
the presence of picrate. Since picrate is a hydroly-
curve that resulted had a small non-zero intercept
sis product of tetryl (Kayser et al. 1984, Kayser
8
Previous Page
