Minefield samples - continued

Table 5. Analytes detected in soils collected above three objects in

an experimental minefield at Sandia National Laboratory.

Soil concentration (g/kg)

2 g:10 mL

25 g:50 mL

Analyte

Rep 1

Rep 2

Rep 1

Rep 2

Salted-in

Cotton swatch: mg quantities of TNT, DNT, and RDX covered in fiberglass window screen

netting

TNT

3.52

3.65

2.85

4.52

1.90

RDX

60.6

48.3

38.2

61.5

33.2

4-Am-DNT

0.95

0.38

5.3

2-Am-DNT

11.4

8.82

6.89

12.6

14.0

Surrogate S-5: HDPE box about the size of anti-tank mine, spray-painted with a mixture of

TNT, DNT, and RDX and containing 75 g of TNT, DNT, and RDX crystals

2,4-DNT

1.55

0.74

0.89

1.49

1.2

TNT

1.94

1.76

0.62

1.63

0.71

RDX

<d

1.47

2.90

1.5

2-Am-DNT

1.07

0.70

0.29

0.64

<d

Land mine AT-3, Bakelite TM62P anti-tank

TNT

1.53

0.96

0.93

1.11

0.67

2-Am-DNT

0.90

<d

samples were preconcentrated and analyzed six

from an experimental minefield at Fort Leonard

times each. These were duplicate 50-mL aliquots

Wood where various defused anti-tank and anti-

of a 5-g/L solution containing the Method 8330

personnel mines are buried. The mines were

analytes plus NG, PETN, and 3,5-DNA. We also

manufactured in the former Yugoslavia and each

preconcentrated blank acetonitrile to test for in-

contained TNT (0.10.2 kg in anti-personnel mines

terfering peaks.

and 5.55.6 kg in anti-tank mines). The first set of

The salted-in soil extracts, preconcentrated by

soil samples was collected two months after the

a factor of 11, yielded concentrations generally con-

mines were buried and revealed which analytes

sistent with what we found in the unconcentrated

are actually present in soils surrounding emplaced

extracts (Table 5). The salting-in procedure did not

land mines. Of the 143 samples we received, only

introduce interfering peaks, nor did we see a

28 had detectable analytes (Table 6) as determined

change in responses for the analytes. The relative

by GC-ECD with confirmation by HPLC-UV. The

standard deviation for six injections of the

most common analytes detected were 2,4-DNT, 4-

preconcentrated standard solutions ranged from

Am-DNT, and 2-Am-DNT, each of which was

a low of 1.1% for NG to a high of 10% for RDX,

found in 27 out of the 28 positive samples and each

at median concentrations greater than 60 g/kg.

and were generally around 5%. Following these

analyses, the injection port liner was removed, and

2,4,6-TNT was found in 24 samples and tended to

white, presumably salt, residue was visible in the

be lower in concentration than the Am-DNTs, in-

liner. With frequent injections of check standards,

dicating that it is undergoing microbiological

the salting-in procedure appears to be a viable

transformation. Median 2,4,6-TNT concentration

was 6.5 g/kg, and the maximum was over 3000

option for preconcentrating samples prior to GC-

g/kg in a soil collected next to a mine. The other

ECD.

Aside from providing a sufficient volume of sol-

two commonly found analytes were 1,3-DNB and

vent for preconcentration, extraction of 25-g soil

2,6-DNT, which had median concentrations of 8.8

and 3.1 g/kg, respectively. Agreement between

samples with 50 mL of solvent did not enhance

detection capability (Table 5) over the standard 2-

duplicates was similar to that observed previously

g soil samples with 10 mL of solvent.

for these analytes (Table 6).

More soils were collected four months after the

Fort Leonard Wood soils

mines were buried; out of 199 samples, 73 had de-

The next sets of soils we received were collected

tectable explosives. Figure 6 shows chromato-

24