in the δ13C value for the soil carbon as TNT
bility of communication between the ground-
water of the upper and lower terrace. Such com-
becomes irreversibly bound to the soil organic
munication could explain the decrease in the δ15N
matter. As discussed above, the change in mass of
values for TNT along transect 2.
carbon in a soil sample resulting from TNT bind-
These results suggest the possibility of using
ing with organic matter in the soil is less than 0.1%.
δ
15N measurements of TNT from groundwater to
This is far too small a change to measure precisely.
The prediction that no changes in the 13C stable
monitor its fate or natural attenuation in the envi-
ronment. Further studies would be required to
isotope ratio would be detected for TNT in solu-
understand how such a measurement could be
tion, even with large changes in concentration of
TNT, was verified. Changes in the δ13C for the soil
used quantitatively. Factors that require investi-
gation are the rate and magnitude of change in
carbon could not be detected as TNT became irre-
the δ15N of TNT versus the concentration gradient
versibly bound to the soil organic matter. In incu-
of TNT across a plume, the contact time with the
bation study 1, the concentration of TNT decreased
soil, and the mass of soil through which the TNT
by six-fold over a very short time, but no signifi-
cant change in the δ13C value for the soil was mea-
has passed. These parameters, as well as organic
and moisture contents, would have to be assessed
sured. In incubation study 2, the concentration of
extractable TNT increased to 25 g/g over 77 days,
for several soil types.
but no significant changes were measured in the
δ13C for soil carbon. Soil sample collected at LAAP
Isotope ratios in soil at LAAP
showed no significant differences between
TNT was detected in most of these samples at
concentrations from 0.1 to 1.8 g/g. Analysis of a
samples from different locations or between sam-
ples from different depths at the same locations.
subset of these samples indicated no difference
between the δ13C values for blank soil samples and
In a positive contrast to the 13C data, the 15N
those that were contaminated with TNT. The
data suggested the feasibility of monitoring natu-
extractable TNT was also analyzed for δ15N; how-
In two field experiments, increasing δ15N values
ever, owing to the low concentration of TNT in
the extracts, reliable measurements could not be
for TNT in the groundwater were observed as the
obtained.
TNT migrated through the soil. Data across three
of the four transects at LAAP showed a signifi-
cant increase in the δ15N values for TNT with
decreases in concentrations. These data are con-
SUMMARY
sistent with the reduction of the nitro groups to
The sensitivities of the δ13C and δ15N analyses
amino functionalities, followed by binding of the
were calculated to be 250 and 750 ng TNT, respec-
amino or di-amino product to the soil carbon
tively. Below 4 ng TNT injected on the column,
(Pennington et al. 1998). Both steps in the process
the δ13C measurement was mass dependent. This
involve reactions of nitrogen groups; therefore,
measuring δ15N values is a very promising moni-
threshold value translated to a soil concentration
of 20 g/g when extracting a 2-g sample with 10
toring tool for natural attenuation of TNT and its
mL of solvent. This is two orders of magnitude
transformation products in groundwater.
higher than the detection limits of the EPA stan-
dard method for explosives (USEPA 1994). The
concentration of TNT in the soil samples collected
CONCLUSIONS
at LAAP ranged from 0.1 to 1.8 g/g. Even where
the soil was in contact with an unlimited source
The objective of this work was to evaluate
whether 13C or 15N stable isotope measurements
of TNT for several weeks, the concentration of
TNT adsorbed to the soil only reached 25 g/g.
of TNT or TNT-contaminated soil could be used
Therefore, a change in the δ13C value for the soil
as a monitoring technique for the natural attenua-
tion of TNT. The 13C data presented here clearly
carbon resulting from the addition of TNT to the
did not support the use of δ13C measurements
soil carbon cannot be measured because of the
because changes in the δ13C value for soil organic
detection limitations of the analysis.
matter were too small to detect. The 15N data,
Sample preconcentration would improve detec-
though, did show that changes in the δ15N values
tion, but may introduce greater measurement
uncertainty. This may address the limitation for
for TNT in groundwater could be detected across
analytes in solution but not for soil. One of the
a contamination plume and suggested the possi-
key objectives was the ability to measure changes
bility of measurable changes being detected in
16
Previous Page
