Minefield samples

Table 4 (cont'd).

Soil concentration (g/g)

Dry

Not

Not

24 hrs.

48 hrs.

air-dried

24 hrs.

48 hrs.

2,3,4-TNT

samples. Rather, it proved yet again that spiking

prior to extraction. The other two replicates were

of soils to simulate field contamination can lead

extracted with 10 mL of acetonitrile. Also, dupli-

to more questions than are answered. During the

cate 25-g aliquots of soils were extracted with 50

subsequent months, we received three batches of

mL of acetonitrile. Following sonication overnight,

soils from minefields, one from Sandia and two

samples were filtered and analyzed by GC-ECD

from Fort Leonard Wood. The soils from Sandia

using the HP-5 separation.

were very dry and not refrigerated. The first batch

Explosives were detected in only those samples

from Fort Leonard Wood was dry and the second

directly above the objects, and with the exception

batch had soils of various moisture contents. We

of RDX above the first object, were close to detec-

took the conservative approach and extracted all

tion limits (Table 5). Mean recoveries for the seven

soils as received.

spiked analytes from the nine matrix spike

samples were 106 7%.

Because the concentrations of the nitroaromatic

Minefield samples

analytes in the unspiked samples were so low, we

Sandia soils and salting-in preconcentration

tested the feasibility of using salting-in

Soils from Sandia were collected from an ex-

preconcentration (Jenkins and Miyares 1991). For

perimental minefield. Three sets of samples were

each object, we preconcentrated the acetonitrile

collected, each set corresponding to a different

from one replicate 25-g sample that was extracted

buried object. The three objects were 1) a cotton

with 50 mL of acetonitrile (Table 5). In a 250-mL

swatch containing milligram quantities of TNT,

glass volumetric flask, 65 g of NaCl (Morton) were

DNT, and RDX covered in fiberglass netting; 2) a

dissolved into 200 mL of reagent-grade water by

high-density polyethylene (HDPE) box about the

stirring with a magnetic stirrer. Once all the salt

size of an anti-tank mine, spray-painted with a

was dissolved, we added 40 mL of acetonitrile soil

mixture of TNT, DNT, and RDX at about 105 g/

extract. All of the acetonitrile dissolved into the

cm2 with about 75 g of TNT, DNT, and RDX crys-

salt water in two out of the three samples. An ad-

ditional 3 mL of acetonitrile was added and stir-

tals inside the box; and 3) a TM-62P anti-tank mine

ring continued. The stirring was stopped and the

(a plastic-cased mine containing 5.7 to 8.3 kg TNT

phases allowed to separate. Then the acetonitrile

[Department of Defense 1999]). Surface soils were

collected directly above each object and 23 and 46

layer (around 3.5 mL) was removed with a glass

cm to the north for a total of nine samples.

Pasteur pipet. 3,4-DNT was added as an internal

standard (1 L of a 200 mg/L solution in acetoni-

When the soils were received in the laboratory,

triplicate 2-g subsamples of each were weighed.

trile). This procedure will result in some deposi-

One of the replicates was spiked with 1 mL of spike

tion of salt in the injection port liner, and we were

solution to yield 50 g/kg of 1,4-DNB, 1,3-DNB,

concerned that after multiple injections this depo-

1,2-DNB, 2,4-DNT, TNT, 4-Am-DNT, and 2-Am-

sition would reduce precision. To determine

DNT; the spiked samples were aged 60 minutes

whether this concern was justified, additional

23