those held at 12C. 1,3,5-TNB accumulated sub-
Table 12. Soilair partition coefficients (Ks/a)
stantially, as well, in soil samples held at 23C,
for TNT-related compounds.
with concentrations ranging from 0.153 to 2.77
Analyte
Soil type
Moist soil
Air-dried soil
mg/kg, depending on the moisture content and
a. At 23C
days held.
1.6 104
3.9 105
2,4,6-TNT
Sand
The accumulation of TNT-related analytes in
2.0 105
Silt
the surface 5 mm of Ft. Edwards Clay was gener-
2.5 106
Clay
ally lower than in the sand and silt (Table 11c).
5.5 103
4.4 104
2,4-DNT
Sand
TNT-related analytes were nearly always below
3.1 104
Silt
detection for air-dried soils, regardless of the tem-
2.8 105
Clay
perature. For the moistened soils, concentrations
2.4 103
1.3 104
1,3-DNB
Sand
ranged from less than detection to 1.57 mg/kg for
2.2 104
Silt
1.9 105
2,4-DNT in the high-moisture soil at 63 days. The
Clay
concentration of 2,4-DNT was higher than that
b. At 4C
for 2,4,6-TNT in nearly every case. Concentra-
3.1 105
1.6 107
2,4,6-TNT
Sand
3.8 105
9.0 106
tions of 1,3-DNB were always less than 1.0 mg/kg
Silt
3.0 107
in the surface of the clay soil.
Clay
Even though the two common environmental
1.5 104
2,4-DNT
Sand
1.3 105
transformation products of TNT (2ADNT and
Silt
1.6 106
Clay
4ADNT) were occasionally detected in these
1.6 104
soils, the accumulation of these analytes over this
1,3-DNB
Sand
7.7 104
several-month experiment was much less than
Silt
6.0 105
Clay
predicted from holding time studies (Grant et al.
c. At 12C
1993, Maskarinec et al. 1991). This may indicate
that analytes moving through the soil in the
2,4,6-TNT
Sand
2.5 106
3.8 105
Silt
vapor state become more tightly associated with
Clay
soil surfaces than what is normally encountered
9.5 104
at higher concentrations at explosives-contami-
2,4-DNT
Sand
6.1 106
Silt
nated sites. At higher concentrations, these com-
Clay
pounds may be present in soil solution to a great-
2.5 104
1,3-DNB
Sand
er extent.
8.5 105
Silt
Clay
Soilair partition coefficients
Since soil concentrations of TNT-related anal-
that 1 mL of headspace vapor contains only one
ytes were obtained immediately after equilib-
thousandth to less than one millionth the mass of
rium headspace measurements were completed,
analyte present as does 1 g of the associated soil.
it is possible to compute soilair partition coeffi-
This suggests that more effort might be made to
cients for each analyte from data obtained in this
exploit the soil's natural ability to preconcentrate
experiment. These partition coefficients are
these vapors, in lieu of, or in addition to, pursu-
important because they compare the amount of
ing more sensitive explosive vapor detectors.
target analyte present in the soil vs. what is
present in air at equilibrium. This information
can be important in deciding whether to concen-
SUMMARY
trate sample acquisition efforts in the air or in the
Experiments were conducted to investigate the
soil to detect mines.
qualitative and quantitative effects of soil barriers
Table 12 shows the effect of soil type, moisture,
at various temperatures on the vapor signature
and temperature on soilair partitioning of explo-
from buried military-grade TNT. Military-grade
sive vapors. Soilair partition coefficients decrease
TNT (110 mg) was buried beneath 2.5 cm of either
with temperature, in keeping with the energetics
a sand, silt, or clay soil in a small glass vial. The
of the sorption process. They are highest for clay
vials were held at temperatures ranging from 12
and lowest for sand, and they are higher for air-
to 23C for times ranging between 63 and 173
dry than moist soil, as suggested by the earlier
days before the soil was removed and analyzed
discussion of headspace concentrations. From a
for explosives signatures. We sampled the head-
detection standpoint, it is worth emphasizing
20
To Contents
Previous Page
