160
140
120
RDX
100
80
60
40
20
0
60
10
20
30
40
50
Time (days)
Figure 5. Mass of RDX released from VS-50 in water at 21.5C.
cients are too large for glassy materials (cf, Berens and
they are essentially a function of partition coefficients
Hopfenberg 1982).
of the ERC between the casings and air (because these
The steady fluxes in air, the initial fluxes on place-
establish the concentration gradients that drive fluxes),
ment in water, and the long-term steady fluxes in water
and their diffusivities in air. Partition coefficients are
are all important for defining potential conditions in
equilibrium constants and, therefore, van't Hoff. Diffus-
the soil environment where mines are ultimately
ivities in air are also exponential in T (Thibodeaux 1996)
deployed. Therefore, we compare these in Table 3 for
and would be similar for all ERC. The similar temper-
the three types of mine for which we have data. We see
ature dependence of DNB, DNT, and TNT flux from
greater differences between water and air for the PMA2
five different mine types is still a bit puzzling. It implies
than the PPM2. This may be caused by differences in
similar sorption energetics for all contaminant and cas-
specific binding, solubilities, and diffusion coefficients
ing combinations studied. The results for other "surro-
of the nitroaromatics among different casing materials
gate" casings were also consistent with this finding
(Leggett and Cragin, in prep.). Similar airwater differ-
(Leggett and Cragin, in prep.).
ences are expected for other plastic materials not tested.
The temperature dependence of vapor flux from cas-
RDX flux was also vastly different in water and air,
ings in a water environment was not studied, but, as
probably a function of its low vapor pressure.
suggested above, fluxes would result from the contam-
inant's diffusivity and solubility in the casing material.
Solubilites (equivalent to equilibrium constants) and
Implications for land mine detection
diffusion coefficients of contaminants in polymers
by vapor sensing
The current results suggest that significant concen-
above their glass transition temperatures are van't Hoff/
trations of ERC would be found in the soil some time
Arrhenius (Romdhane et al. 1995, Aminabhavi et al.
after mines are buried. This has recently been confirmed
1996, Xiao et al. 1997). This appears to be true below
in the analysis of experimental minefield soils at Fort
their glass transition temperatures as well, but with a
Leonard Wood (Jenkins et al. 2000). Direct compari-
reduced dependence on T (Romdhane et al. 1995). It
son of lab flux and field soil concentrations of ERC is
may be worth noting that two of the materials used in
impossible because temperature fluctuation and envi-
the mine casings studied (PMA1A, PMA2, and TMA5)
ronmental fate and transport processes alter their spa-
were supposedly below their glass transition tempera-
tures, polyvinylchloride (81C) and polystyrene
tial and temporal distribution. A simple example will
(100C). However, their actual glass transition tempera-
illustrate the consistency of results, though. Analysis
of soil samples around a TMA5 antitank mine after
tures may have been lowered by plasticization. ERC
4 months of burial revealed concentrations of less than
diffusion coefficients estimated from the magnitude of
1 to 800 ng of DNT/g of soil (Jenkins et al., in press).
the observed fluxes and plastic solubilities (Leggett and
Taking the mean flux at 20C from Table 2 and averag-
Cragin, in prep.) suggest that all the casings were plas-
ing it over the estimated surface area of this mine, 2700
ticized in these experiments. That is, the diffusion coeffi-
8
Previous Page
